LITERATURE REVIEW: UNDERSTANDING THE CURRENT STATE …€¦ · literature review: understanding the current state of autonomous technologies to improve ... report code: smruc-ogp-2015-015

LITERATURE REVIEW:

UNDERSTANDING THE CURRENT STATE OF AUTONOMOUS TECHNOLOGIES TO

IMPROVE/EXPAND OBSERVATION AND DETECTION OF MARINE SPECIES

Title:AutonomousTechnology

DATE:July2016

REPORTCODE:SMRUC-OGP-2015-015

THISREPORTISTOBECITEDAS:VERFUSSU.K.,ANICETO,A.S.,BIUW,M.,FIELDING,S.GILLESPIE,D.,HARRIS,

D.,JIMENEZ,G.,JOHNSTON,P.,PLUNKETT,R.,SIVERTSEN,A.,SOLBØ,A.,STORVOLD,R.ANDWYATT,R.2016.

LITERATURE REVIEW: UNDERSTANDING THE CURRENT STATE OF AUTONOMOUS TECHNOLOGIES TO

IMPROVE/EXPANDOBSERVATIONANDDETECTIONOFMARINESPECIES.REPORTNUMBERSMRUC-OGP2015-

015PROVIDEDTOIOGP,July2016.

Coverphoto:AutoNautcompletingaPAMmission.©PeterBromley.

For its part, the Buyer acknowledges that Reports supplied by the Seller as part of the Servicesmay be

misleadingifnotreadintheirentirety,andcanmisrepresentthepositionifpresentedinselectivelyedited

form.Accordingly,theBuyerundertakesthatitwillmakeuseofReportsonlyinuneditedform,andwilluse

reasonableendeavourstoprocurethatitsclientundertheMainContractdoeslikewise.Asaminimum,afull

copyofourReportmustbeappendedtothebroaderReporttotheclient.

3


DATE:July2016


1 Contents 1 Contents........................................................................................................................................................3

2 Figures...........................................................................................................................................................7

3 Tables............................................................................................................................................................8

4 ExecutiveSummary.....................................................................................................................................11

4.1 Objectives...........................................................................................................................................11

4.2 Approach............................................................................................................................................12

4.3 Outcome.............................................................................................................................................13

4.3.1 UAS.................................................................................................................................................13

4.3.2 AUVandASV..................................................................................................................................14

4.3.3 Futurework....................................................................................................................................16

5 Quickguideforprospectiveusers...............................................................................................................18

6 Projectframeworkandapproach................................................................................................................22

7 Introduction.................................................................................................................................................24

7.1 Monitoringtypes................................................................................................................................25

7.1.1 Mitigationmonitoring....................................................................................................................25

7.1.2 Animalpopulationsurveys.............................................................................................................26

7.1.3 Focal-follows...................................................................................................................................27

7.2 Autonomousplatformtypes...............................................................................................................27

7.2.1 Poweredaircraft(UAS)...................................................................................................................27

7.2.2 Motorizedgliders(UAS).................................................................................................................28

7.2.3 Kites(UAS)......................................................................................................................................28

7.2.4 Lighter-than-airaircraftsystems(UAS)..........................................................................................29

7.2.5 Propellerdrivenunderwatercraft(AUV).......................................................................................30

7.2.6 Buoyancygliders(AUV)..................................................................................................................31

7.2.7 Poweredsurfacecrafts(ASV).........................................................................................................33

7.2.8 Self-poweredsurfacevehicles(ASV)..............................................................................................35

4


DATE:July2016


7.2.9 Drifter(ASVandAUV).....................................................................................................................36

7.3 Sensortypes........................................................................................................................................37

7.3.1 Electro-opticalimagingsensors......................................................................................................37

7.3.2 PAMsystems..................................................................................................................................39

7.3.3 AAMsensors...................................................................................................................................42

7.3.4 Animal-bornetags..........................................................................................................................44

8 Evaluationofautonomoussystems.............................................................................................................44

8.1 Criteriaandmetrics............................................................................................................................44

8.1.1 Collectionofsurveydata................................................................................................................45


8.1.3 Operationalaspects........................................................................................................................47

8.1.4 Datarelay.......................................................................................................................................48

8.1.5 Furtherinformationgathered........................................................................................................49

8.2 Evaluationresults...............................................................................................................................50

8.2.1 UASplatforms.................................................................................................................................50

8.2.2 UAS-sensors....................................................................................................................................55

8.2.3 AUV/ASVplatforms......................................................................................................................57

8.2.4 AUV/ASV-sensors..........................................................................................................................61

8.3 Discussion...........................................................................................................................................65

8.3.1 How to use the evaluation results to find the appropriate autonomous platform / sensor

combination................................................................................................................................................65

8.3.2 Strengthandweaknessesofautonomousvehicletypes...............................................................71

8.3.3 Platformtypesformitigationmonitoring......................................................................................73

8.3.4 Platformtypesforpopulationmonitoringandfocal-follows.........................................................75

8.3.5 Technologiesthatmightbeusedinfuturefieldtrials....................................................................77

8.3.6 Datagapsandrecommendations...................................................................................................79

9 Furtherinformation.....................................................................................................................................86

9.1 Requirementsofautonomousvehiclesformarineanimalmonitoring..............................................86

5


DATE:July2016



9.1.2 Animalpopulationmonitoring.......................................................................................................89

9.1.3 Focalanimalstudies.......................................................................................................................92

9.2 Currentknowledgeonautonomousvehicles.....................................................................................93

9.2.1 Independentsystems.....................................................................................................................93

9.2.1 Cooperativesystems....................................................................................................................116

9.3 Operationalaspectstoconsider.......................................................................................................117

9.3.1 TechnicalReliability......................................................................................................................118

9.3.2 MissionDuration–Powerconsiderations....................................................................................119

9.3.3 Logistics........................................................................................................................................119

9.3.4 RemoteOperation........................................................................................................................119

9.3.5 Interactionwithothermarine/airspaceusers..............................................................................120

9.4 Regulatory/politicalbarriers...........................................................................................................120

9.4.1 Regulationsfortheairspace.........................................................................................................120

9.4.2 Regulationsforthesea.................................................................................................................122

9.5 Technicalchallengesofmanaging,storingandanalysinglargeamountofdata..............................124

9.5.1 Electro-opticalimagingsensors....................................................................................................126

9.5.2 PAMsystems................................................................................................................................127

9.5.3 AAMsensors.................................................................................................................................129

10 Acknowledgements...............................................................................................................................130

11 Appendix...............................................................................................................................................131

11.1 Listofcriteriaandmetricsforplatform,sensoranddatarelay.......................................................131

11.1.1 Platform...................................................................................................................................131

11.1.2 Sensor......................................................................................................................................132

11.1.3 Datarelay.................................................................................................................................134

11.2 Shortlistofplatformsandsensors....................................................................................................135

11.2.1 Poweredaircrafts.....................................................................................................................135

6


DATE:July2016


11.2.2 Kites..........................................................................................................................................137

11.2.3 Lighter-than-airaircraftsystems..............................................................................................137

11.2.4 Propellerdrivenunderwatercraftandpoweredsurfacecraft................................................138

11.2.5 Autonomousunderwaterbuoyancygliders.............................................................................143

11.2.6 Self-poweredsurfacevehiclesanddriftingsensorpackages...................................................145

11.2.7 PAM..........................................................................................................................................146

11.2.8 AAM.........................................................................................................................................148

11.3 Comparisonmatrices........................................................................................................................151

11.3.1 Listofplatforms.......................................................................................................................151

11.3.2 Platformtechnicaldimensions.................................................................................................155

11.3.3 Platformothertechnicaldetails...............................................................................................163

11.3.4 Platformcosts..........................................................................................................................167

11.3.5 Platformsurveycapabilities.....................................................................................................171

11.3.6 Platformoperationalinformation............................................................................................176

11.3.7 Platformmanningrequirements..............................................................................................184

11.3.8 Listofsensor............................................................................................................................188

11.3.9 Sensortechnicaldetails............................................................................................................190

11.3.10 Sensorcosts.............................................................................................................................193

11.3.11 SensorsurveycapabilitiesI......................................................................................................195

11.3.12 SensorsurveycapabilitiesII.....................................................................................................198

11.3.13 Sensoroperationalrequirements............................................................................................200

11.3.14 Sensormanningrequirements.................................................................................................203

11.3.15 Sensorfurtherinformation......................................................................................................204

11.3.16 ListofDataRelaysystems........................................................................................................207

11.3.17 Datarelaytechnicaldetails......................................................................................................209

11.4 Evaluationmatrices..........................................................................................................................211

11.4.1 UASsensorversusplatformmatrix..........................................................................................211

7


DATE:July2016


11.4.2 ASV/AUVsensorversusplatformmatrix.................................................................................212

11.5 Suppliercontacts..............................................................................................................................216

11.6 Surveyingwildlifepopulations..........................................................................................................220

11.6.1 Estimatingabsolutepopulationdensityandabundance–anoverview.................................220

11.6.2 Methodstoestimatedetectionprobability.............................................................................222

12 GlossaryofTerms,AcronymsandAbbreviations..................................................................................225

13 Referencelist........................................................................................................................................228

2 Figures FIGURE1.DECISIONTOOLFORFINDINGTHEAPPROPRIATEPLATFORM/SENSORCOMBINATIONFORASPECIFICMONITORINGTASKAND

ANIMALTYPE.THEMONITORINGTYPESAREDIVIDEDINTOMITIGATIONMONITORINGINAREASEITHERCLEARFROMORBUSY

WITHOTHEROPERATIONALGEARORTRAFFIC,POPULATIONMONITORINGINTHESHORT-TERM(HOURS,DAYS)ORLONG-TERM

(WEEKS,MONTHS)ANDFOCAL-FOLLOWSCONDUCTEDWITHSTATICORMOBILESYSTEMS.ANSWERINGQUESTIONSQ1(WHICH

MONITORINGTYPESHOULDBECONDUCTED)WITHDIAGRAMAANDQ2(WHICHTYPEOFANIMALNEEDSTOBEMONITORED)

WITHDIAGRAMBWILLHIGHLIGHTTHEMOSTIMPORTANTSYSTEMREQUIREMENTSANDTHEANIMALBEHAVIOURPOTENTIALLY

TRIGGERINGDETECTION,LEADINGTOTHESENSOR/PLATFORMCOMBINATIONBESTSUITEDFORASPECIFICMONITORINGTASK.

......................................................................................................................................................................21

FIGURE2.EXAMPLESFORUAS:LEFTPICTURE:SKYPROWLERUAS©KROSSBLADEAEROSPACE,RIGHTPICTURE:LANDINGOF

CRYOWINGUAS.PHOTOBYNORUT..................................................................................................................28

FIGURE3.KITESUSI62TAKE-OFF.PHOTOBYTHAMM,2011..........................................................................................29

FIGURE4.THELIGHTER-THAN-AIRCRAFTSYSTEM"BLIMP-CAM"ANDCOMPONENTSNECESSARYFORFLIGHTS.PHOTOBYHODGSON,

2007..............................................................................................................................................................30

FIGURE5.PROPELLERAUVREMUS100INAFIELDSTUDY(LEFT,BYANREYSHCHERBINA,WHOI)ANDDETAILOFTHEMODEL

(RIGHT,KONGSBERGMARITIMEWEBSITE)..............................................................................................................31

FIGURE6.UNDERWATERBUOYANCYGLIDERSSEAGLIDER(LEFT)ANDSLOCUM(RIGHT)(IMAGES©DAMIENGUIHEN)..................32

FIGURE7.ANAUVBOATTYPEVIKNESDESIGNEDANDDEVELOPEDBYTHENORWEGIANUNIVERSITYOFSCIENCEANDTECHNOLOGYIN

CONJUNCTIONWITHMARINEROBOTICS(IMAGE©NORWEGIANUNIVERSITYOFSCIENCEANDTECHNOLOGY)....................33

FIGURE8.CATAMARANS:LEFTPICTURE:SPRINGER,ACATAMARANBASEDASVDEVELOPEDBYPLYMOUTHUNIVERSITY(IMAGE©

PLYMOUTHUNIVERSITY),RIGHTPICTURE:ASVC-ENDURO,ALARGESCALECATAMARANBASEDASV(IMAGE©ASV).........34

FIGURE9.THESEMI-SUBMERSIBLEASV-6300DESIGNEDBYC&CTECHNOLOGIESANDASV(IMAGEFROMWOLKING,2011)......35

8


DATE:July2016


FIGURE10.UNDERWATERPHOTOGRAPHOFALIQUIDROBOTICSWAVEGLIDERSHOWINGTHESUBUNIT,WHOSEMOVEABLEFINS

CONVERTVERTICALMOTIONFROMWAVESINTOFORWARDPOWERANDTHESURFACEFLOAT(PHOTO,LIQUIDROBOTICS)......36

FIGURE11.THEUKNATIONALOCEANOGRAPHYCENTRESAUTONOMOUSVEHICLEFLEET......................................................96

FIGURE12.SUNFISHTRACKINGPROJECT(PINTO,2014)................................................................................................117

FIGURE13.CURRENTCOMMERCIALSMALLUASRULESINCOUNTRIESWITHSTRONGANDGROWINGMARKETSFORUASSYSTEMS

(BONGGAY,2016)..........................................................................................................................................122

FIGURE14.SCHEMATICDIAGRAMSOFFOURDIFFERENTSENSORDATAHANDLINGCOMBINATIONS.THESEAREINTENDEDPRIMARILYAS

EXAMPLESANDMANYMORECOMBINATIONSEXIST.A)SIMPLESENSINGANDDATASTORAGE,B)SENSOR,BASICPROCESSING

ANDDATASTORAGE,C)REAL-TIMETRANSMISSIONOFRAWDATAWITHPROCESSINGONSHORE,D)ONBOARDPROCESSINGTO

REDUCEDATAVOLUMEPRIORTOTRANSMISSIONOFAMINIMALAMOUNTOFDATA......................................................125

FIGURE15.EXAMPLESOFREALANDFALSEWHISTLEDETECTIONSBOTHWITHOUTANDWITHFULLSPECTROGRAMINFORMATION.A)A

FALSEDETECTIONCAUSEDBYTHEHYDRAULICSOFANEARBYENGINE.B)ADOLPHINWHISTLE.........................................129

3 Tables TABLE1.COMMONDESIGNFEATURESOFELECTRICBUOYANCYGLIDERS(FROMDAVISETAL.,2002ANDWOODANDMIERZWA,

2013)............................................................................................................................................................59

TABLE2.EVALUATIONMATRIXOFDETECTION/CLASSIFICATIONCAPABILITIESOFSENSORS(INCLUDEDINTHISREVIEW)CAPABLEOF

REAL-TIMEDETECTION.WITHTHEHELPOFTHESESENSORS,ONEISCAPABLEOFDETECTING(DETECT)ORNOTCAPABLEOF

DETECTING(NO)MARINEANIMALS.SOMEOFTHESENSORSDELIVERDATAWITHWHICHMARINEANIMALSCANBECLASSIFIED

(TOSOMEEXTENT)TOSPECIESORSPECIESGROUPS(CLASSIFY)...................................................................................68

TABLE3.PROPERTIESGENERALLYAPPLYINGFORAUTONOMOUSVEHICLESFORTHEDATARELAYSYSTEMSSATELLITELINKSAND

WIRELESSSYSTEMS............................................................................................................................................69

TABLE4.EVALUATIONMATRIXOFMONITORINGCAPABILITIESOFSENSOR/PLATFORMTYPES.GIVENTHEYMEETTHECRITERIAAS

OUTLINEDINSECTION8,THESPECIFICSENSOR/PLATFORMTYPESLISTEDAREINGENERALWELLSUITED(�),SUITED(X)ORNOT

APPLICABLE(-)FORTHECERTAINMONITORINGTYPES(EXCEPTIONSMAYBEPOSSIBLE).THEMONITORINGTYPESARE

FURTHERMOREDIVIDEDINTOMITIGATIONMONITORINGINAREASEITHERCLEARFROMORBUSYWITHOTHEROPERATIONALGEAR

ORTRAFFIC,SHORT-TERM(HOURS,DAYS)ORLONG-TERM(WEEKS,MONTHS)POPULATIONMONITORINGANDFOCAL-FOLLOWS

CONDUCTEDWITHSTATICORMOBILESYSTEMS.L-T-A=LIGHTER-THAN-AIRAIRCRAFT....................................................70

TABLE5.RECOMMENDEDTECHNOLOGYTYPESTHATMIGHTBEUSEDINFUTUREFIELDTRIALSFORTHEDIFFERENTMONITORINGTYPES.

THEMONITORINGTYPESAREFURTHERMOREDIVIDEDINTOMITIGATIONMONITORINGINAREASEITHERCLEARFROMORBUSY

WITHOTHEROPERATIONALGEARORTRAFFIC,SHORT-TERM(HOURS,DAYS)ORLONG-TERM(WEEKS,MONTHS)MONITORING

ANDFOCAL-FOLLOWSCONDUCTEDWITHSTATICORMOBILESYSTEMS.L-T-A=LIGHTER-THAN-AIRAIRCRAFT.......................79

9


DATE:July2016


TABLE6.SHORTLISTOFRECOMMENDATIONSWITHMEDIUM(M)ORHIGH(H)PRIORITYANDWITHMEDIUM(M)ORHIGH(H)

URGENCY,ALONGWITHTHERELATEDTOPICANDDATAGAP.......................................................................................84

TABLE7:LISTOFPUBLISHEDSTUDIESINVOLVINGUASWITHAPPLICATIONINTHEMARINEENVIRONMENT................................104

TABLE8.LISTOFSTUDIESINVOLVINGAUVANDASVWITHPOSSIBLEAPPLICATIONINTHEMARINEENVIRONMENT.....................107

TABLE9.DATAVOLUME(GIGABYTES)FORDIFFERENTRECORDINGBANDWIDTHSANDRECORDINGDURATIONSFORASINGLE

HYDROPHONERECORDEDAT16BITACCURACY......................................................................................................128

TABLE10TECHNOLOGYREADINESSLEVELSUSEDINTHEEVALUATIONMATRICES*...............................................................134

TABLE11.HAZARDCLASSLEVELSRELATEDTOBATTERYORFUELTYPEOFANEVALUATEDSYSTEM.OILSOROTHERMATERIALUSEDIN

MANUFACTURINGOFSYSTEMSTHATMAYHAVEHAZARDSARENOTCONSIDERED..........................................................135

TABLE12.AUTONOMOUSPLATFORMSINCLUDEDINTHISREVIEW,THEIRSYSTEMCLASS(POWEREDANDUNPOWEREDASV,

PROPELLERANDGLIDERAUV,LIGHTER-THAN-AIRAIRCRAFTS(L-T-A)UASS,KITESANDPOWEREDFIXEDWINGUASS),SYSTEM

NAME,MANUFACTURERAND/ORDESIGNER/ORIGINATINGUNIVERSITYASWELLASTHEIRTECHNOLOGYREADINESSLEVELAS

DEFINEDINTABLE10.NOTETHATCELLSWITHTHESAMEKINDOFINFORMATIONFORSYSTEMSOFTHESAMEMANUFACTURER

MAYBEMERGED.............................................................................................................................................151

TABLE13.TECHNICALDIMENSIONSOFTHEAUTONOMOUSPLATFORMSLISTEDINTABLE12.GIVENARETHEIRMISSIONDURATION

ANDFACTORSLIMITINGTHEMISSIONDURATION,THEIRMINIMUM,MAXIMUMANDTYPICALSPEED,THEIRVERTICAL

OPERATIONALRANGE,THEIRENVIRONMENTALPERFORMANCELIMITS.THEIROPERATIONALSIZEANDWEIGHTASWELLAS

PACKING/FREIGHTWEIGHTANDSIZEAREGIVENANDTHERANGETHEYCANBECONTROLLEDWITHIN.NOTETHATCELLSWITHTHE

SAMEKINDOFINFORMATIONFORSYSTEMSOFTHESAMEMANUFACTURERMAYBEMERGED.ABBREVIATIONS:L-T-A:LIGHTER-

THAN-AIRAIRCRAFTS,NA:NOTAPPLICABLE,N.P.:NOTPROVIDED,LWH:LENGTH,WIDTH,HEIGHT.*UNLESSOTHERWISE

SPECIFIED.......................................................................................................................................................155

TABLE14.OTHERTECHNICALDETAILSOFTHEAUTONOMOUSPLATFORMSLISTEDINTABLE12.NOTETHATCELLSWITHTHESAMEKIND

OFINFORMATIONFORSYSTEMSOFTHESAMEMANUFACTURERMAYBEMERGED.ABBREVIATIONS:L-T-A:LIGHTER-THAN-AIR

AIRCRAFTS,NA:NOTAPPLICABLE,N.P.:NOTPROVIDED,SSS:SIDESCANSONAR,CTD:CONDUCTIVITY,TEMPERATURE,DEPTH,

IR:INFRARED,SVTP:SOUND,VELOCITY,TEMPERATUREANDTEMPERATURE,SUNA:SUBMERSIBLEULTRAVIOLETNITRATE

ANALYZER,LND,COTS:COMMERCIALOFF-THE-SHELF,PAR:PHOTOSYNTHETICALLYACTIVERADIATION,RINKO:

PHOSPHORESCENTDOSENSOR..........................................................................................................................163

TABLE15.COSTDETAILSOFTHEAUTONOMOUSPLATFORMSLISTEDINTABLE12.GIVENISTHEPLATFORMPRICEFORPURCHASEOR

RENTAL,THECOSTSFOROPERATION,MAINTENANCE,BATTERY/FUELCOSTSANDCOSTSFORPILOTINGTELEMETRY.NOTETHAT

CELLSWITHTHESAMEKINDOFINFORMATIONFORSYSTEMSOFTHESAMEMANUFACTURERMAYBEMERGED.PRICESAREFOR

VEHICLESONLYWITHOUTSENSORS.ADDITIONOFSENSORPACKAGESCANSIGNIFICANTLYINCREASETHEPRICE.ABBREVIATIONS:

L-T-A:LIGHTER-THAN-AIRAIRCRAFTS,NA:NOTAPPLICABLE,N.P.:NOTPROVIDED.......................................................167

10


DATE:July2016


TABLE16.SURVEYCAPABILITIESOFTHEAUTONOMOUSPLATFORMSLISTEDINTABLE12.GIVENISTHECAPABILITYOFTHEPLATFORMS

FORTRACKSETTING,INCLUDINGITSIMPLEMENTATIONDETAILSANDLIMITATIONS,THEIRABILITYFORSTATIONKEEPING,

AUTONOMOUSDECISIONMAKINGANDCLOCKSYNCHRONISATIONINCL.THEIRLIMITATIONS.NOTETHATCELLSWITHTHESAME

KINDOFINFORMATIONFORSYSTEMSOFTHESAMEMANUFACTURERMAYBEMERGED.ABBREVIATIONS:L-T-A:LIGHTER-THAN-

AIRAIRCRAFTS,NA:NOTAPPLICABLE,N.P.:NOTPROVIDED.....................................................................................171

TABLE17.OPERATIONALINFORMATIONONTHEAUTONOMOUSPLATFORMSLISTEDINTABLE12.GIVENISTHEFUELTYPE,FAILSAFE

MODETYPE,TRANSPONDER,DETECTANDAVOIDCAPACITY,GROUNDSTATIONINTERFACE,ITSLEVELOFAUTONOMY,THE

PAYLOADCAPACITYINTERMSOFPOWER,SPACEANDWEIGHTASWELLASTHEDEPLOYMENTANDRECOVERYPROCEDURE.NOTE

THATCELLSWITHTHESAMEKINDOFINFORMATIONFORSYSTEMSOFTHESAMEMANUFACTURERMAYBEMERGED.

ABBREVIATIONS:L-T-A:LIGHTER-THAN-AIRAIRCRAFTS,NA:NOTAPPLICABLE,N.P.:NOTPROVIDED................................176

TABLE18.MANNINGREQUIREMENTSFORTHEAUTONOMOUSPLATFORMSLISTEDINTABLE12.GIVENISXX.NOTETHATCELLSWITH

THESAMEKINDOFINFORMATIONFORSYSTEMSOFTHESAMEMANUFACTURERMAYBEMERGED.ABBREVIATIONS:L-T-A:

LIGHTER-THAN-AIRAIRCRAFTS,NA:NOTAPPLICABLE,N.P.:NOTPROVIDED................................................................184

TABLE19.SENSORSTHATMAYBEINTEGRATEDINTOAUTONOMOUSVEHICLESANDINCLUDEDINTHISREVIEW,THEIRSYSTEMCLASS

(AAM,PAM,VIDEO),SYSTEMNAME,MANUFACTURERASWELLASTHEIRTECHNOLOGYREADINESSLEVELASDEFINEDINTABLE

10).NOTETHATCELLSWITHTHESAMEKINDOFINFORMATIONFORSYSTEMSOFTHESAMEMANUFACTURERMAYBEMERGED.

ABBREVIATIONS:AAM:ACTIVEACOUSTICMONITORING,PAM:PASSIVEACOUSTICMONITORING...................................188

TABLE20.TECHNICALDETAILSOFTHESENSORSLISTEDINTABLE19.GIVENARETHEENVIRONMENTALPERFORMANCELIMITSOFTHE

SENSORS,THEIROPERATIONALASWELLASPACKING/FREIGHTSIZEANDWEIGHT,HAZARDCLASSASDEFINEDINTABLE11,ANY

FACTORSINTERFERINGWITHTHEOPERATIONOFTHESENSOR,ANDTHESENSOR’STYPEOFINTERFACE.NOTETHATCELLSWITH

THESAMEKINDOFINFORMATIONFORSYSTEMSOFTHESAMEMANUFACTURERMAYBEMERGED.ABBREVIATIONS:AAM:

ACTIVEACOUSTICMONITORING,PAM:PASSIVEACOUSTICMONITORING,NA:NOTAPPLICABLE,N.P.:NOTPROVIDED........190

TABLE21.COSTSSENSORSLISTEDINTABLE19.GIVENARETHEPURCHASEANDRENTALPRICE,ASWELLASOPERATIONAL,

MAINTENANCE,BATTERY/FUELASWELLASINTEGRATIONCOSTS.NOTETHATCELLSWITHTHESAMEKINDOFINFORMATIONFOR

SYSTEMSOFTHESAMEMANUFACTURERMAYBEMERGED.ABBREVIATIONS:AAM:ACTIVEACOUSTICMONITORING,PAM:

PASSIVEACOUSTICMONITORING,NA:NOTAPPLICABLE,N.P.:NOTPROVIDED.............................................................193

TABLE22.FIRSTSETOFSURVEYCAPABILITYCRITERIAFORSENSORSLISTEDINTABLE19.GIVENISTHEDATATYPECOLLECTED,DATA

PROCESSINGDETAILS,IFSPECIESIDENTIFICATIONISPOSSIBLEANDIFSO,WHICHTYPE,ANDIFTHEBEARING,DIRECTAND/OR

HORIZONTALRANGETOTHEANIMALISESTIMABLEWITHTHESENSORDATA.NOTETHATCELLSWITHTHESAMEKINDOF

INFORMATIONFORSYSTEMSOFTHESAMEMANUFACTURERMAYBEMERGED.ABBREVIATIONS:AAM:ACTIVEACOUSTIC

MONITORING,PAM:PASSIVEACOUSTICMONITORING,NA:NOTAPPLICABLE,N.P.:NOTPROVIDED................................195

TABLE23.SECONDSETOFSURVEYCAPABILITYCRITERIAFORSENSORSLISTEDINTABLE19.GIVENISTHEREAL-TIMEDATA

TRANSMISSIONCAPABILITIESINCL.DETAILSWATISTRANSMITTED,THEABILITYFORCLOCKSYNCHRONISATIONANDITS

LIMITATIONSANDFORACOUSTICSENSORS,IFRECEIVINGLEVELSOFRECEIVEDSIGNALSISESTIMABLEASWELLASAMBIENTNOISE

11


DATE:July2016


LEVELS,ANDIFCTDDATAAREOBTAINEDSIMULTANEOUSLY.NOTETHATCELLSWITHTHESAMEKINDOFINFORMATIONFOR

SYSTEMSOFTHESAMEMANUFACTURERMAYBEMERGED.ABBREVIATIONS:AAM:ACTIVEACOUSTICMONITORING,PAM:

PASSIVEACOUSTICMONITORING,NA:NOTAPPLICABLE,N.P.:NOTPROVIDED.............................................................198

TABLE24.OPERATIONALREQUIREMENTSOFTHESENSORSLISTEDINTABLE19.GIVENISTHEIRLEVELOFAUTONOMY,DEPLOYMENT

ANDRECOVERYPROCEDURE,THEIRPAYLOADPOWER,ANDINTERNALASWELLASEXTERNALPAYLOADCAPACITYINTERMSOF

SPACEANDWEIGHT.NOTETHATCELLSWITHTHESAMEKINDOFINFORMATIONFORSYSTEMSOFTHESAMEMANUFACTURER

MAYBEMERGED.ABBREVIATIONS:AAM:ACTIVEACOUSTICMONITORING,PAM:PASSIVEACOUSTICMONITORING,NA:NOT

APPLICABLE,N.P.:NOTPROVIDED......................................................................................................................200

TABLE25.MANNINGREQUIREMENTSFORTHESENSORSLISTEDINTABLE19.GIVENISTHEMINIMUMNUMBEROFPEOPLEFOR

OPERATION,THEVESSELREQUIREMENTS,THEMANNINGREQUIREMENTSFORDEPLOYMENTANDRECOVERYASWELLASTHE

TRAININGNEEDSFOROPERATINGTHESENSOR.NOTETHATCELLSWITHTHESAMEKINDOFINFORMATIONFORSYSTEMSOFTHE

SAMEMANUFACTURERMAYBEMERGED.ABBREVIATIONS:AAM:ACTIVEACOUSTICMONITORING,PAM:PASSIVEACOUSTIC

MONITORING,NA:NOTAPPLICABLE,N.P.:NOTPROVIDED......................................................................................203

TABLE26.FURTHERINFORMATIONONTHESENSORSLISTEDINTABLE19.GIVENISTHERAWDATAVOLUME,THECAPABILITIESTO

REDUCEREAL-TIMEDATA,LINKSANDCOMMENTS.ABBREVIATIONS:AAM:ACTIVEACOUSTICMONITORING,PAM:PASSIVE

ACOUSTICMONITORING,NA:NOTAPPLICABLE,N.P.:NOTPROVIDED........................................................................204

TABLE27.DATARELAYSYSTEMSTHATCANBEUSEDFORAUTONOMOUSVEHICLES..............................................................207

TABLE28.TECHNICALDETAILSOFTHEDATARELAYSYSTEMS:THEIROPERATIONALSIZEANDWEIGHT,POWER,BANDWIDTH,RANGE,

PURCHASEANDOPERATIONALCOSTSASWELLASUSAGERESTRICTIONS,DATADELAYANDFURTHERCOMMENTS.................209

TABLE29.COMPILATIONOFWHICHVIDEOSENSORCOULDPOTENTIALLY(O)ORDEFINITIVELYBECOMBINED(X),WHICHCANNOTBE

COMBINED(-),ANDWHICHCOMBINATIONWILLWORKBUTWITHLIMITEDMISSIONDURATION(L).GREENMARKEDCELLS

INDICATESENSOR/PLATFORMCOMBINATIONSTHATAREALREADYINTEGRATED.NA:NOTAPPLICABLE..............................211

TABLE30.COMPILATIONOFWHICHPAMORAAMSENSORCOULDPOTENTIALLYORDEFINITIVELYBECOMBINED,SENSOR/PLATFORM

COMBINATIONSTHATAREALREADYINTEGRATED,WHICHCOMBINATIONSCANNOTBECOMBINEDANDWHICHCOMBINATION

WILLWORKBUTWITHLIMITEDMISSIONDURATION................................................................................................212

TABLE31.LISTOFSUPPLIERCONTACTSOFTHEPLATFORMSANDSENSORSINCLUDEDINTHISREVIEW.......................................216

4 Executive Summary 4.1 Objectives

Thisreporthasbeenconductedtoenhancetheunderstandingofhowautonomousvehiclesystemscanbeused

tomonitormarinemammalsandothermarineanimals intheframeof industrialactivitiesoftheoilandgas

(O&G) industry. The purpose of suchmonitoringwould be formitigationmeasures, to estimate population

12


DATE:July2016


statusandtrendsinareasofinterest,ortoconductfine-scalebehavoiuralstudiestoevaluatepotentialimpacts

ofanthropogenicinducednoiseontheseanimals.Keyobjectivesofthereportinclude:

• An exhaustive review and evaluation of available platform and sensor technologies that are most

pertinenttooilandgasoperationsformarineanimalmonitoring;

• Information on the key data types that can be obtained, addressing the technical challenges of

managing,storingandanalysingthelargeamountofdatagatheredbytheautonomoussystems;and

• Recommendations of platform technologies, sensors and data characteristics that would be most

relevantfortheO&Gindustry.

4.2 Approach

Inordertoaddresstheobjectives,thereportprovidesacomprehensiveevaluationofthestatusandpotential

ofautonomousaerialandmarineplatformsandsensortechnologiesthataremostpertinenttoO&Goperations

for conductingmitigationmonitoring, population surveys and/or trackingofmarine animals such asmarine

mammals,seaturtlesandfish.Themainplatformsandsensorsconsideredinthisrevieware:

• UnmannedAerialSystems(UAS)includingpoweredaircraft,gliders,kitesandlighter-than-airaircraft;

• UASsensorsincludingthermalIR,Non-thermalIR,RGBandvideocameras;

• Autonomous Underwater Vehicles (AUV) and Autonomous Surface Vehicles (ASV) including

autonomouspropellerdrivencraft,autonomousunderwaterbuoyancygliders,autonomouspowered

surfacecraft,selfpoweredsurfacevehiclesanddriftingsensorpackages;

• AUV/ASVsensorsincludingPassiveAcousticMonitoring(PAM)andActiveAcousticMonitoring(AAM)

sensorsaswellasanimal-bornetranspondertags.

Evaluation criteria and metrics were defined, which were used to evaluate the autonomous systems’

applicabilityandsuitabilityforvariousmonitoringactivities.Furthercriteriaallowedtheevaluationofpotential

platformandsensorcombinationsandconsiderations tobe taken intoaccountduringprojectplanning.The

informationgatheredwascompiledintomatricestoallowacomparisonofthedifferentsystems’operational

capabilities,dataprocessingandtransferaspectsandtheirpotentialformitigationmonitoringandsurveydata

collection. Further informationwas compiled and summarised toprovide further insights into the following

topics:

• Requirementsofautonomousvehiclesformarineanimalmonitoring;

• Currentknowledgeaboutautonomousvehicles;

• Operationalaspectstoconsider;

• Regulatory/politicalbarriers;

13


DATE:July2016


• Technicalchallengesofmanaging,storingandanalysinglargeamountofdata.

4.3 Outcome

The report illustrates that it is not possible to give simple ‘one-size-fits-all’ suggestions for choosing the

appropriateautonomousvehicletoconductdifferentkindsofmonitoringapplications,astherearemanyproject

dependentfactorsthatneedtobeconsidered.Thelargevarietyoftargetspeciesrequirestheselectionofthe

appropriatesensortype/platformtypecombination(ratherthanaspecificsystem).Thesame is truefor the

differentmonitoringtypes(mitigationmonitoring,densitymonitoring,focal-follow),combinedwiththechoice

ofdatarelaysystem.Largerandmorecomplexsystemsaregenerallymorecapable,butarelikelytobemore

expensiveandrequiregreatersupportinginfrastructure.Thetechnicalandoperationaldetailsofanautonomous

platformwillneedtobetailoredtothespecificneedsoftheexplorationandproduction(E&P)project,e.g.the

areaofinterest,thepropertiesoftheplatformthattheautonomousvehiclewilldeployedfrom,therequired

surveylength,projectbudgetandmore.Herewewillgiveanoverviewofwhichplatformtypesarebesttouse

for variousmonitoring typesby comparingeach classofplatformandhighlighting theirparticular strengths

relatingtothespecificmonitoringtypes.

OncethedetailsofanE&Pprojectareknown,thecomparisonandevaluationmatricesinthisreportwillthen

helptoselecttheappropriateplatform/sensorcombination,andthediscussionsectionwillhelptounderstand

howwellthatsystemwilladdressthetaskinhand.Theestablishmentofacomputer-basedexpertsystembased

onthisreviewandtheprecedingreviewonlowvisibilitymonitoringtechniquesaspublishedinVerfussetal.,

(2016)wouldbeahelpfulandappropriatetooltohandlethevastamountofinformationgathered.Thisexpert

systemshouldbebasedontheevaluationcriteriaestablishedinbothreviews,andshouldprovidealistofthe

mostappropriatesystemsandmethodsasanoutputbasedonthemonitoringrequirementsfedintotheexpert

system.Theestablishmentofsuchatoolwas,however,outsidethescopeofthisreview.

4.3.1 UAS

UAShaverecentlybeenintroducedasalternativeplatformstoovercomesomeofthelimitationsofmanned-

aerialsurveys,mainlyfortheirabilitytoperformthesametasksmoreefficiently,safely,andatafractionofthe

cost(NeiningerandHacker,2011).Recentdevelopmentshaveturnedpresent-dayUASintorealisticalternatives

tomannedsystems,suchaslongerflightdurations,improvedmissionsafety,flightrepeatabilityduetoimproved

autopilotsystems,andreducedoperationalcostswhencomparedtomannedaircraft.Thepotentialadvantages

ofanunmannedplatform,however,dependonfactorssuchasaircraftflightcapability,sensortype,mission

objective,andcurrentregulatoryrequirementsforoperationsoftheparticularplatform(Wattsetal.,2010a).In

recent years, UAS have been integrated into a variety of field studies involving a range of species and

applications.However,therehasbeenlittlefocusontheapplicationofthesesystemswithamoreO&Gindustrial

approach,concentratingontheirbenefitsfore.g.monitoringformitigationpurposesandtheirintegrationwith

other autonomous technologies (e.g. coordinationof unmannedaerial, surface, andunderwater vehicles to

14


DATE:July2016


createacommunicationnetworkthatiscapableofsimultaneouslysurveyingthesameareausingavarietyof

sensorsinordertomaximisethechancesofdetectionandfordatacross-validationbetweenplatforms).

The variety of UAS configurations currently available is a great advantage for offshoremonitoring. Smaller

platformsallowcoverageofsmallerareasata lowcost(e.g.exclusionzonesduringseismicoperations),and

largerplatformsallowcoverageoflargerareas,providingvaluableinformationbothbefore,during,andafter

offshore operations. Hence, it is possible to cover a wide range of projects using different UAS models.

Unfortunately,thecomplexityofthelargersystemsrequiresmultipleexperiencedpersonneltooperatethem

andmanyofthesystemsarecurrentlyonlyavailableformilitaryuse.Largerplatformsmayalsorequirerunways

similartothoserequiredbymanned-airplanes,whichlimitsthepossibilityfordeploymentsatsea.Conversely,

theperformanceofsmallerplatformsmaybe limitedbyenvironmentalconditionsandpayloadsare limited,

meaningthattheymayonlybeabletocarrylesssophisticatedsensorequipmentand/oroperateforshorter

periodsoftime.

OfthethreeclassesofUASplatformsconsideredinthisreview,poweredaircrafthavemanyofthecapabilities

requiredforaerialsurveysofmarineanimals.Kitesandlighter-than-airaircraft,suchasballoons,arecost-saving

alternatives tosomefixedandrotarywingsystems, thoughtheyrelyheavilyonweatherconditionsandare

difficulttocontrolwhenfollowingapredefinedtrack.Intermsofindividualfocalfollows,however,lighter-than-

airaircraftshouldbeconsideredasapotentialcandidateplatform,thoughtheirabilitytofollowanimalswillbe

restrictedtothemanoeuvrabilityofthesupportvesseltowhichtheyaretethered.Poweredaircraftaregood

candidateplatformsforindividual-basedmonitoringastheycanbepilotedtofollowanimals(thoughfixed-wing

systemsmaystruggletohoverabovestationary,orslowmoving,focalanimals),whilekiteshavelimitedcontrol.

UAS are also suitable formitigationmonitoring provided they have real-time detection options.With long

operationalranges,theyareespeciallysuitedformonitoringtheareaaheadofaseismicvesselinordertodetect

marineanimalswithinanexclusionzonearoundtheanticipatedstartlocationofthesoundsource,assuggested

byVerfussetal.(2016).

ThisstudymakesthefollowingshortlistofrequirementsforaUASmonitoringsystem:stabilisedhigh-resolution

videofordetection,highresolutionimagesforidentification,real-timegeocodingofimagedataandacomputer

systemforefficientanalysisofrecordeddata.Ifthedataaretobetransmittedtotheground,asneededfor

mitigationmonitoring,ahighbandwidthdatalinkisalsorequired.

4.3.2 AUVandASV

AutonomousUnderwaterandSurfaceVehicles(AUVandASV)aredividedinthisreportintofivemaincategories:

propeller driven underwater craft, underwater buoyancy gliders, powered surface vehicles, self-powered

surfacevehiclesanddrifters.Thevehiclesrangefromrelativelysmallsize,whichcanbeliftedbyoneortwo

personsanddeployedfromshoreorasmall inflatable,tolargediesel-poweredsurfacevesselswithbespoke

launchandrecoverysystems(Griffiths,2002).Manysurfaceandunderwaterautonomousvehiclesarenowused

15


DATE:July2016


on a regular basis by research institutions world-wide. Primarily, they are used for the collection of

oceanographicdata(e.g.seasurfaceorwatercolumntemperatureandsalinityforinstance)whilethedetection

of largerorganisms,suchasmarinemammalsandfish,generallyremainsanicheareaofresearch.Powered

underwaterandsurfacevehicleshavetheadvantageofgreatercontrol,butgenerallyhavelimiteddeployment

times(oftenhourstodays).Buoyancydrivensystemsaregenerallysmallenoughtobedeployedmanuallyfrom

asmallvesselandarecapableofmulti-monthmissions,givingthemgreatpotentialforlongtermpopulation

monitoring.Wavedrivensystemstendtobeslightlylargerbutcanstilloftenbedeployedandrecoveredusing

modestfacilities.Thelowestcostsystemsaredrifters,whichhavenocontroloverwheretheygo,butcouldbe

acost-effectivewayofcollectinglargedatasamples.

Avarietyofactiveandpassiveacousticsensorsareavailable,whicheitherhavebeen,orhavethepotentialto

be, integratedintotheseplatforms.Whilenoteverysensor issuitableforeveryplatform,foreachplatform,

thereisatleastonesensorwhichmightbeincorporatedintoitandvice-versa.AnumberofPassiveAcoustic

Monitoring (PAM) sensors are able to run automatic detection algorithms to automatically pick outmarine

mammal calls. However, these generally need checking by a human operator before they can be used in

managementdecisionmaking,soconsiderablequantitiesofPAMdatamusteitherbestoredonthedevicefor

futureanalysisortransmittedtoanoperatoriftheyarerequiredinreal-time.Solutionsexistwhichcantransport

sufficientdataovershortrangesforreal-timeoperationevenforhighfrequencydata.Mostsystemsreviewed

hereonlydeployasinglehydrophoneandareincapableofdirectlymeasuringbearingsorrangestodetected

sounds.

AutomaticprocessingforActiveAcousticMonitoring(AAM)islessadvancedthanforPAM,withdatagenerally

being stored for future analysis. As with PAM, it should be possible to transmit sufficient data over short

distancesforreal-timedecisionmaking.Itisalsopossibletoattachacoustictagstotargetanimalstoidentify

andtrackindividualsinrealtimewithareceivingPAMsystemonanASVorAUV.Obviously,akeyrequirement

isthattheanimalistaggedinthefirstplace.

ThemainrestrictionsgoverningtheuseofAUVandASVformitigationmonitoringinthevicinityofseismicsurvey

vesselsareoperational.Thedeploymentofanautonomousvehicleclosetoanoperationalseismicsurveyvessel

carriestheriskthatthevehiclemayinterferewithseismicoperationsandtheindustrywouldneedtobewilling

toacceptthisriskwhendeployinganautonomoussurfaceorunderwatervehicleaheadofthesurveyvessel.It

isthereforeunlikelythatAUVandASVwillhavearoleinmonitoringaroundseismicvesselsinthenearfuture.

Ontheplusside,seismicvesselshavemanycompetenttechnicalpersonnelonboardandthepotentialtodeploy,

serviceandoperateautonomousplatforms;however,dependingonthecircumstancestheremaynotberoom

onboardeitherforadditionalequipmentorpersonnelrequiredtooperateautonomousplatforms.AUVand

ASVmayhavearoletoplaymonitoringfurtheraheadofseismicvesselsasanaidtoforwardplanning.Thismay

be important in sensitiveareaswherePAMon theASV /AUVcouldenhancedetectioncapabilityofbaleen

whalesandpromptmitigationmonitoringby,forinstance,thermalimaging.Additionally,ASVandAUVmaybe

16


DATE:July2016


important when mitigating around static sources, such as rig operations where space for equipment and

personnelmaybelimited.Thismightbeofparticularinterestwithregardtotheuseofexplosives(suchasfor

well-headdecommissioningorremovalofunexplodedordnanceUXO)duetotheclearrequirementtoremove

personnelfromthevicinity.

Thelongmissiondurationsofferedbyself-poweredsurfacevehiclesandbuoyancyglidersmakesthesesystems

particularly appealing for long term population monitoring. The choice of vehicle would depend on the

characteristics of the area to be surveyed (e.g. size, water depth and current strength). For population

monitoring, data need not be transmitted in real time, but can be archived to storagemedia. A particular

advantageoflowpowervehiclesforPAMisthelownoiseproducedbythesystemitself,andmoreimportantly,

theconsistencyofnoisebetweenplatforms.Themainlimitationstotheuseofthistechnologyforpopulation

monitoring are the same as for vessel based acoustic surveys, namely the need to understand detection

probabilityfordifferentspeciesasafunctionofrangefromthedetectorandtheneedforrequiredbehavioural

data(e.g.groupsizesorcueproductionrates)sothatcountsofanimaldetectionscanbetranslatedintoanimal

densityestimates.Thelowernoiseandmissiondurationcapabilitiesoflowpowersurfacevesselsalsosuggest

anadditional future role,not inmitigation,but inunderwatersoundmeasurement forverificationof sound

levelsemittedfromsoundsourcesintheframeofE&Pactivities.

4.3.3 Futurework

Autonomoussystemshavethepotentialtodetectawidevarietyofspeciesindifferentoperationalscenarios.

Theprimarylimitationsontheuseofautonomoustechnologiesduringmitigationinthevicinityofseismicvessels

are operational. If there is a desire on the part of industry to overcome these problems, then suitable

engineeringandenvironmentalscienceteamsboth,insideandoutsideofindustry,willneedtofurtherassess

andfindsolutionstothesequitesignificantproblems.Forpopulationmonitoring,technologiesexistwhichcan

collect data for awide rangeof species anddevelopment shouldbe tailored to specific species groups and

operationalareassincethiswillguidethechoiceoftechnology.Forspecificscenarios,thecostandbenefitof

using autonomous vehicles should also be comparedwithmore traditionalmethodsof data collection (e.g.

vesseloraerialvisualsurvey).

Wheremultipletechnologieshavethepotentialtodetectaparticularspecies,werecommendthatside-by-side

comparisons should be conducted between current monitoring methods (e.g. visual vessel based or aerial

surveys) and different autonomous solutions. This is valuable for evaluating the degree to which the data

acquired by these technologies can be complementary. Particularly in the case of PAM, simultaneous

deploymentsofvarioussystemscanhelptoestimatethedetectionprobabilitiesofagivensystembyusingthe

otherplatformstoconduct“trial-based”detectionprobabilityestimation.Thiswould,however,demandahigh

degreeofcontrolovertheconditionsinwhichtheequipmentisdeployedandastudymayneedtorunforsome

timetoobtainstatisticallyusefulsamplesizes;nevertheless,itwouldberequiredtofullyunderstandtherelative

capabilitiesofallthesesystems.

17


DATE:July2016


Real-time decisionmaking and analysis of large data sets inevitably require the use of automatic detection

algorithmsforallsensortypes.Automaticdetectionbothreducestheamountofdatathatneedtobechecked

byahumanoperator,andmakesitmorepracticaltosendthedatathroughlowerbandwidthdatarelaysystems.

The development of detection algorithms is further advanced for PAM than for video and AAM, in most

situations.Therefore,AAMandvideofutureresearchshouldfocusondevelopingdetectionalgorithmsanddata

compression techniques for returning summarised and/or raw data over low bandwidth communications.

However,whiledetectionandclassificationalgorithmswillalwayscontinuetoimprove,itmustbeunderstood

thattheywillneverbeperfectandwillcontinuetorequireahumanoperatoratsomepointintheprocessing

chain.This fundamental limitationshouldbe taken intoaccountaswemove forwardwithdeveloping these

systems.While we would not suggest that further research into algorithms for automatic detection is not

worthwhile, research into both the magnitude and the effects of mis-detection andmis-classification, and

investmentintosystemsthataidhumanobserversinthedecisionprocessesisalsoofhighimportance.

Finally,behaviouralinformationonmanymarineanimalspeciesformsahugedatagap.Auxiliarybehavioural

data relating to the study species (e.g. group size, dive duration or call production rate) are required for

abundanceestimationand/ormitigation(inbothcaseswhereunderstandingprobabilityofdetectioniscrucial).

Therefore,continuedbehaviouralstudiesofallspeciesofinterestwillbeimportantforsuccessfulsurveyingand

monitoring.

18


DATE:July2016


5 Quick guide for prospective users Anyone wishing to use the information provided in this review to find an appropriate platform/sensor

combination should start by answering the two questions below to narrow down the list of candidate

technologies(seealsoFigure1).Beneatheachquestion,wepointtothesections,figuresandtablesthatcontain

backgroundinformation,theimportantcriteriaandmetricsaswellaswheretofindinformationtoaddressthe

most importantpointsthatneedtobeconsideredwhendecidingontheappropriateplatformorsensor.An

overviewofthestrengthsandweaknessesofthedifferentplatformsisgiveninsection8.3.2.

When themost suitableplatformand sensor typehasbeen identified, anda short-listof the systemsmost

suitable for the specific monitoring task is specified, the possible platform / sensor combination is mainly

determinedbythepayloadcapacityoftheplatform(Table17)andtheoperationalsizeandweightofthesensor

(Table20).Anoverviewofpotentialplatform/sensorcombinationsisgiveninTable29andTable30.Thedata

relayclassandsystemthatshouldbeusedcanbedeterminedbyreadingsection8.1.4andconsultingTable27

andTable28,bearinginmindthatoperationalsizeandweightneedtofitthepayloadoftheplatform.

Fromthisfinalshort listofplatforms,sensorsanddatarelaysystems,thefinaldecisionisdependentonthe

users demand and limitations, e.g. limited manning requirements (manning requirements information for

platforms=>Table18, forsensors=>Table25)or limitedbudget(costsofplatforms=>Table15,sensors=>

Table21,relaysystems=>Table28).Oncetheplatform,sensoranddatarelaysystemshavebeenselected,itis

advisabletocontactthecorrespondingsuppliers(=>Table31)toconfirmthesuitabilityofthechosensystem

forthedemandedtasksandtoretrievepotentialupdatesontheinformationgatheredinthisreview,asthisis

arapidlydevelopingfield.

Q1:Whatkindofmonitoringshouldbeconducted?

a) Mitigationmonitoring

• Mostimportantpointstoconsider:

o Real-timedetectionof theanimalof interest isobligatory, thereforeonlysensorswith

real-timeoptionsareappropriate(=>Table23)

o Ahighdetectionlikelihoodoftheanimalisrequired,whichinfluencesthechoiceofthe

sensor(=>Q2).Concurrentuseofdifferentmethodsincreasesthedetectionprobability

o Operational period and range need to be sufficient tomeet the temporal and spatial

requirementsofthemonitoring(=>Table13)

o Manoeuvrability is important to keepwithin themonitoring area (=> Table16),which

excludesdriftersfortheuseofmitigationmonitoring

19


DATE:July2016


o Inoperationallybusyareas,AUVandASVarenotrecommendedduetotheriskofcollision

andentanglementwithoperationalgear

o Synchronisedoperationofa fleetwidensthefieldofviewandallowstocovera larger

area(section9.2.1)

• Requirements:section9.1.1

• Whichplatformtypetouse:section8.3.3,Figure1,Table4

• Criteriaandmetricstoconsider:section8.1.2

• Backgroundinformation:section7.1.1

b) Populationsurvey


o Operational period, speed and range need to be sufficient tomeet the temporal and

spatialrequirementsofthesurvey(=>Table13),whichexcludesmostpoweredsystems

forlong-termmonitoring

o Trackkeepingisimportant,atleastforshort-termsurveys(=>Table16)





c) Focal-animalstudy


o Trackingofthefocalanimalisrequired.Systemchoicedependsonthesensorcapabilities

toestimatebearingandrangetotheanimalduringstationaryfocalfollows(i.e.animalis

trackedwithinsensor’sview)(=>Table22)andthetrackingabilityoftheplatformduring

mobilefocalfollows(i.e.platformfollowsanimal)(=>Table16)

o Manoeuvrabilityisimportantduringmobilefocalfollows(=>Table16)

o Operational period, speed and range need to be sufficient tomeet the temporal and

spatialrequirements(=>Table13)



20


DATE:July2016




Q2:Whatkindofanimalneedstobemonitored?

a) Speciesregularlyseenatorneartheseasurface

• Sensor:Electro-opticalimagingsensors(section7.3.1)

• Animal types: Whales, dolphins/porpoises, seals, turtles, large solitary fish (e.g. basking

sharks),swarmfish(specifiedforsensorsystemsinTable22)

• Detectionprobability:increaseswithtimespentatornearseasurface

• Mostsuitableplatform(s)=UAS

b) Speciesinthewatercolumn

• Sensor:AAMsensors(section7.3.3)



• Detectionprobability:increaseswithsizeofanimaloranimalschool

• Mostsuitableplatform(s)=AUV/ASV

c) Specieswithfrequentanddistinctivevocalisations

• Sensor:PAMsensors(section7.3.2)

• Animaltypes:Whales,dolphins/porpoises(specifiedforsensorsystemsinTable22)

• Detectionprobability:increaseswithincreasedvocalisationrateand/orvocalisationduration

• Mostsuitableplatform(s)=self-poweredAUV/ASV,poweredAUV/ASVwithlowself-noise

d) Animalssuitablefortagging

• Sensor:Animal-borntranspondingacoustictags incombinationwithahydrophone(section

7.3.4)



• Note:animalneedstobeactivelytagged;onlysuitableforfocal-animalstudies

• Mostsuitableplatform(s)=self-poweredAUV/ASV,poweredAUV/ASVwithlowself-noise

21


DATE:July2016


Figure1.Decisiontoolforfindingtheappropriateplatform/sensorcombinationforaspecificmonitoringtaskandanimaltype.ThemonitoringtypesaredividedintoMitigationmonitoringinareaseitherclearfromorbusywithotheroperationalgearor traffic,Populationmonitoring in the short-term(hours,days)or long-term(weeks,months)andFocal-followsconductedwithstaticormobile systems.AnsweringquestionsQ1 (whichmonitoring typeshouldbeconducted)withdiagramAandQ2(whichtypeofanimalneedstobemonitored)withdiagramBwillhighlightthemostimportantsystemrequirementsandtheanimalbehaviourpotentiallytriggeringdetection, leadingtothesensor/platformcombinationbestsuitedforaspecificmonitoringtask.

Whales,dolphins,porpoises

Frequentdistinct

vocalisation

Frequentlynearoratsea

surface

Passiveacousticmonitoringsensors

Animaltype

Animalstatus/behaviour

Suitablesensortype

Mostsuitableplatformtype

Seals,turtles,fish

Yes

Yes

Yes

No

Yes

Belowseasurface

Electro-opticalimagingsensors

Activeacousticmonitoringsensors

No

Self-poweredAUV/ASVPoweredlowself-noise

AUV/ASV

PoweredUAS,Kites,Lighter-than-airaircrafts

Self-poweredAUV/ASVPoweredAUV/ASV

Suitablefortagging

Yes

No

Animal-borntransponding tags

Self-poweredAUV/ASVPoweredlowself-noise

AUV/ASV

Mitigation Area

Clear

Busy

Length

Short-term

Range Static

Mobile

PoweredUAS,kites,lighter-than-airUAS,poweredandself-

poweredASV/AUV

PoweredUAS,kites,lighter-than-airUAS

PoweredUAS,(kites,lighter-than-airUAS),poweredandself-poweredASV/AUV,Drifters

Self-poweredASV/AUV,Drifters

Poweredandself-poweredASV/AUV,Animalbornetags,(poweredUAS,kites,lighter-

than-airUAS)

PoweredUAS,(kites,lighter-than-airUAS),poweredandself-

poweredASV/AUV

Real-timedetection,

Highdetectionprobability

Reliablevehicle

navigation

Speedampleforsurveycompletion

Longmissionduration

Trackingcapability

Stationkeepingcapability

Long-term

Monitoringtype

Monitoringconditions

Suitableunmannedvehicles

Systemrequirements

Population

Focal-animal

Yes

Yes

Yes

No

No

A

B

22


DATE:July2016


6 Project framework and approach This report was funded by the International Association of Oil & Gas Producers following the Request for

Proposals (RFP)Number JIP III-15-02“AutonomousAerialandMarineTechnologyUnderstanding -Literature

review on understanding the current state of autonomous technologies to improve/expand observation and

detection ofmarine species” from the Joint Industry Programmeon E&P Sound andMarine Life - Phase III,

releasedon24thFebruary2015.

TheobjectivesoutlinedintheRFPweretoenhancetheunderstandingofhowautonomousvehiclesystemscan

beusedformitigationandimpactmonitoringandforestimatingpopulationstatusandtrends.

KeyobjectivesoftheRFPincluded:

• An exhaustive review of available UAS and AUV platforms and sensors technologies that aremost

pertinenttooilandgasoperationsformonitoringofmarinespecies,includingdetailsoncostofsystems

andsensors,currentstateof technologymaturation, sensorperformance,system lifespan, limits to

operating conditions (weather, depth, currents, etc.), alongwith relevant references to field tests,

sourcesofciteddataandkeypointsofcontactforeachsystem.

• Key data types (e.g. digital photographs, digital sound files, oceanographic data, e.g. temperature,

salinity,depth,currents,bathymetry,GPSlocations,etc.)andthedata’spotentialusesforassessingthe

effectsofoilandgasindustryexplorationandproduction(E&P)aswellasgeophysicaloperations(e.g.

seismic surveys) on marine species. Technical challenges of managing, storing and analysing large

amountofdataproducedbytheautonomoussystemsshouldbeaddressed.

• Recommendations of platform technologies, sensors and data characteristics that would be most

relevantfortheE&Pindustrytopursue,eitherviafurtheradvancementoftechnologydevelopmentor

fieldtrialsofexistingpromisingsystems, includingcombinationsofsensors/platformsbestsuitedto

particularspecies,geographic,andoperationalconditions.

SMRUConsulting teamed upwith the SeaMammal ResearchUnit (SMRU), Akvaplan-niva AS, theNorthern

Research Institute (NORUT), theCentre for Research into Ecological andEnvironmentalModelling (CREEM),

SeicheLtdandBritishAntarcticSurvey(BAS)toformateamofhighlyexperiencedexpertsinthefieldofmarine

autonomoustechnologiesandmarineanimalmonitoringtoundertakeacomprehensivereviewonthestatus

and potential of aerial and marine autonomous technologies for conducting marine animal mitigation

monitoringand/orpopulationsurveysaswellasmonitoringanimalmovementandbehaviouralreactions.

Theprojectteamconductedanextensiveliteraturereviewandinformationsearchcombinedwithcontacting

suppliers and developerswhere the relevant informationwas not publically available. The literature search

focusedonspecificinformationasrequestedbytheRFPandoutlinedinthisreport,andcriteriaimportantfor

evaluatingkeyfeaturesofthesystems,suchas

23


DATE:July2016


• platformapplicabilityduringE&Poperations,

• operationalrequirements,

• dataprocessingandtransfercapabilities,

• platformabilityto

o conductmitigationmonitoring,

o collectappropriatesurveydataforanimalpopulationmonitoring,

o monitoranimalmovementandbehaviouralreactions.

Asubsequentevaluationofthegatheredmaterialprovidedareviewofthestateoftheartofautonomousaerial

andmarineplatformsandsensortechnologies.Operationalissuesanddataprocessingandtransfercapabilities

were reviewed and a comprehensive evaluation of the systems for impact monitoring and survey data

acquisitionwas undertaken. This review identified knowledge gaps and areas for further development and

researchleadingtorecommendationsonfuturestudies.

Themainplatformsandsensorsconsideredwere:

• UnmannedAerialSystems(UAS)includingpoweredaircraft,gliders,kitesandlighter-than-airaircraft

(e.g.balloons);

• UASsensorsincludingthermalIR,Non-thermalIR,RGBandvideocameras;

• Autonomous Underwater Vehicles (AUV) and Autonomous Surface Vehicles (ASV) including

autonomouspropellerdrivencraft,autonomousunderwaterbuoyancygliders,autonomouspowered

surfacecraft,selfpoweredsurfacevehiclesanddriftingsensorpackages;

• AUV/ASVsensorsincludingPassiveAcousticMonitoring(PAM)andActiveAcousticMonitoring(AAM)

sensorsaswellasanimal-bornetags.

Thereportprovidesashortintroductionintothedifferentmonitoringtypesconsideredinthisreview(section

7.1)aswellastheconsideredautonomousplatforms(section7.2)andsensortypes(section7.3).Thereport

thendefinesandexplainsthedifferentcriteriaandmetricsfortheevaluationofthesystems(section7.1),gives

theresultsoftheevaluation(section7.2)anddiscussesthese(section7.3).

Inadditiontotheevaluationof thedifferentsystems, further informationwascompiledandsummarisedto

provideinsightsintothefollowingtopics:

• Requirementsofautonomousvehiclesformarineanimalmonitoring(section8.1);

• Currentknowledgeonautonomousvehicles(section8.2);

• Operationalaspectstoconsider(section8.3);

24


DATE:July2016


• Regulatory/politicalbarriers(section8.4);

• Technicalchallengesofmanaging,storingandanalysinglargeamountofdata(section8.5).

Theappendixcontains:

• Tableswiththedefinitionsofthecriteriaandmetrics(section10.1);

• Shortlistsofplatformsandsensors(section10.2);

• Comparison(section10.3)andevaluation(section10.4)matrices;

• Suppliercontacts(section10.5);

• Anexplanatorysectiononsurveyingwildlifepopulations(section10.6).

7 Introduction Industrial activities that involve pile driving (such as the construction of offshorewind farms or oil-rigs) or

activationof soundsources (suchasseismicsurveys)often involve the introductionofanthropogenicsound

energyintothewater.Concernhasgrownthatunderwatersoundhasthepotentialtoimpactmarinemammals

and othermarine animals, whichmay result in auditory injury and/or behavioural changes (Gordon, 2003;

Ketten,1995;Lucke,2009;Pirotta,2014).Thereisademandtoincreasemonitoringeffortsinordertocontribute

tobaselinedatafortheregionofinterestaswellastoassessandminimiseanthropogenicimpactonmarine

animals.Methodsarerequiredthatcaneffectivelybeusedforreal-timeornearreal-timemonitoringfortimely

decision making during operational environmental risk management and permit compliance (mitigation

monitoring), formonitoring individual animal's response to direct impacts (focal-follows), and for exploring

populationsizeanddistribution(populationmonitoring).Soundmayinfluencemarineanimalsatrangesthat

cannotbemonitoredfromtheimmediatevicinityofanoperationalsite.Theuseofautonomousvehiclesfor

mitigationmonitoring,populationmonitoringorfocal-followsthereforerepresentpotentialadvantagesforsuch

applications.Suchvehiclescanbedeployedintheair(UAS)orinthewater(AUV,ASV),andarethereforesuitable

forthedetectionofmarineanimalsattheseasurfaceandunderwater;Vehicles’applicationsreducehealthand

safetyriskstohumanoperators;andtheycanoperateatlongrangesfarbeyondtheanimaldetectionrangesof

humanobservers.UASplatformsincludepoweredaircraft,gliders,kitesandlighter-than-airaircraft,whichwill

mostlybeusedwithsensorssuchasthermal infra-red(IR),non-thermal IR,RedGreenBlue(RGB)andvideo

cameras.AUVsincludeunderwaterbuoyancyglidersandpropellerdrivencraft,butexcluderemotelyoperated

vehicles(ROV),whicharetetheredunderwatercraft,fromourdefinitionofAUV.ASVincludepoweredsurface

craftaswellasself-poweredsurfacevehiclesanddriftingsensorpackages.ForAUVandASVtherelevantsensors

formarineanimalmonitoringwithregardstoE&PactivityincludePassiveAcousticMonitoring(PAM)andActive

AcousticMonitoring(AAM)sensorsaswellasanimal-bornetags.IR,RGBandvideocamerascanbedeployed

25


DATE:July2016


onAUV andASV, however are not useful for the kind ofmonitoring considered in this reviewdue to their

extremelylowunderwatervisiblerange(unlessinveryclearwaters).

Inthefollowing,thedifferenttypesofmonitoringmethodsrelevantformarineanimalmonitoringwithregards

toE&Pactivitiesaswellasthedifferenttypesofplatformsandsensorsaredescribedtoprovidebackground

informationonthetopicsdiscussedwithinthisreview.

7.1 Monitoringtypes

7.1.1 Mitigationmonitoring

Mitigationmonitoringisoftenrequiredwhereanthropogenicactivitiesinvolvetheemissionofsoundthatmay

impactmarinevertebrates.Itaimsforatimelydetectionofmarineanimalswithinapre-definedareaarounda

soundsourcetoallowformitigationmeasurestobetaken.Boththeanimalspeciesandthesizeoftheareato

bemonitoreddependontheregulationsthataretobemetbythespecificoperation.Forexample,guidelines

fortheimplementationofmarinemammalmitigationduringanthropogenicactivitiessuchasseismicsurveys

are often country specific.Most guidelines requiremonitoring for allmarinemammals, though sometimes

monitoringrequirementsareonlyinplaceforspecificsub-groups(e.g.baleenwhales,largertoothedwhales,

cetaceans)orspecificspecies.Seaturtles,polarbearsandwalrusmayalsobeofconcern.Theareathatrequires

mitigationmonitoringrangesfrom500mto3,000mandbeyond,andalsodependsontheregulations.This

areamayincludetheexclusionzoneonly.Theexclusionzoneistheareawithinwhichmitigationmeasureshave

to be taken upon an animal sighting. Some regulations request themonitoring of a larger area around the

exclusionzoneforensuringatimelydetectionofananimalbeforeitenterstheexclusionzone.Thetotalareato

bemonitored(includingtheexclusionzone)ishereafterreferredtoasthemonitoringzone.Anotherpointto

consideristhedurationofthemitigationmonitoring.Monitoringneedstobestartedaminimumof30minutes

beforetheactualoperationstarts.Someguidelinesspecifyarequirementformonitoringof60minutesor120

minutesbeforetheoperationstarts,dependingonthesituationorspeciesofconcern.Foracomprehensive

reviewoftheguidelinespleaseseeVerfussetal.(2016).

Marine mammal (and other animal) mitigation monitoring is traditionally performed by Marine Mammal

Observers (MMOs)orProtectedSpeciesObservers (PSO)visuallyscanningtheseasurfaceof themonitoring

zonetodetecttargetspecies.TheMMO/PSOsarelocatedeitherontheoperatingvesseloronsatellitevessels

accompanyingtheoperation.Toincreasetheprobabilityofdetectingsomespecies,visualmonitoringisoften

accompanied by acoustic monitoring with PAM systems. Thermal imaging systems have also been used,

especiallyforthedetectionofanimalsinsituationsoflowvisibility,whenMMO/PSOswillnotbeabletoconduct

visualmonitoringe.g.duetolackoflight.Imagingsystemstodatehavebeenoperatedfromvessels.

26


DATE:July2016


7.1.2 Animalpopulationsurveys

Surveyscanaimtoanswerawiderangeofquestionsaboutapopulationofinterest.Inthisreport,wefocuson

a few (linked) applications of relevance to theO&G industry: estimating absolute population abundance or

density, assessing the spatial distribution of populations and how this may change over time and, finally,

assessingbehaviouralresponsestoanthropogenicimpacts.Theseapplicationsarelinkedbecausespatialand/or

temporalchangesinthesizeand/ordistributionofapopulation(whichmaybecausedbyanthropogenicfactors)

are typically best quantified by investigating changes in absolute abundance or density. Therefore, survey

requirementstoestimateabsoluteabundanceordensitywillbeinitiallyconsideredbeforeoutliningwhatwould

alsoberequiredtodetectspatialandtemporalchangesinthesemetricsoveranareaofinterestforaspecies

ofinterest,andidentifywhetheranydetectedchangesareduetoanthropogenicimpacts.

Marinemammalpopulationsizeshavebeentraditionallyestimatedusingsightingsdatacollectedfromvessel-

oraerial-basedsurveys(e.g.Garneretal.,1999).Inrecentyears,passiveacousticdatahavebeenincreasingly

collected fromboth vessel-towed instruments and static instruments (either fixed to a surface buoy or the

seafloor),whichhavebeenusedtoestimateanimaldensityorabundance(Marquesetal.,2013b).Approaches

to estimate fish population sizes include (1) conductingdedicated stock assessment surveys, (2)monitoring

catch-per-unit-effortdataor(3)taggingindividualswithactiveacoustictagsthatcanbedetectedusingpassive

acoustic receivers (Hilborn andWalters, 1992). However, the latter abundance estimationmethods rely on

catchingfish.Activeacousticmonitoringusingechosounders(SimmondsandMacLennan,2005)isanalternative

approachforusingdistancesamplingmethodologythatdoesnotrequireanimalstobecaughtandthecollected

datacanreadilybeanalysedinasimilarframeworktovisualandpassiveacousticsurveydata(e.g.seeCoxet

al.,2011forakrillexample).Largerfish,suchaswhalesharksandtuna,canalsobesurveyedfromaircraft(e.g.

Rowatetal.,2009;Bauer,2015).Turtlesarecommonlyvisuallysurveyedfromvesselsandaircraft(e.g.Benson

etal.,2007;Bresetteetal.,2010),thoughland-basedcountsatnestingbeachesandin-watermethodsusing

snorkellingobserversalsoexist(e.g.Weberetal.,2013;Stadleretal.,2014).Inrecentyears,theuseofactive

acousticstomonitorturtleshasbeeninvestigated.Turtleshavebeenfittedwithactiveacoustictransmitting

tags(Thumsetal.,2013)andtheabilitytouseechosounderstodetectturtleshasalsobeenrecentlyexplored

(Pérez-Arjona,2013).

Whethercollectingsightingsoracoustic(eitheractiveorpassive)dataduringwildlifesurveys,theaimisoften

toestimatethetotalnumberofanimalsinagivenareabasedonsomeformofencounterwithananimal,or

groupofanimalsandthemovementsoftheseanimals.Numbersofvisualoracousticencountersaresometimes

presentedasindicesofrelativeabundanceordensity.However,inordertolinkanyspatialortemporalpatterns

presentinsuchindicestochangesintheunderlyingabundanceordistributionofanimals,itmustbeassumed

thatanequalproportionofanimalsremainundetectedacrossspaceandtime.Thisisaverystrongassumption

tomakeand,therefore,statisticaldataanalysismethodshavebeendevelopedthatcorrectobserveddatafor

27


DATE:July2016


undetectedanimals,leadingtoabsoluteestimatesofanimalabundanceordensity(Borchersetal.,2002)(see

section11.6.1fordetailedinformation).

Surveystypicallytakeplacealongplannedsurveytransectsorfromstaticmonitoringpoints.Insomesurveys,it

maybepossibletodetectallanimalswithinthetransectlinesorpoints.Suchsurveytypesareknownasstrip

transectsampling(usingtransectlines)orplotsampling(usingpoints).Whenanimalsaresuspectedtobemissed

within surveyedareas, thereare severalmethods that canbeused toestimate theprobabilityofdetection

(detailedinsection11.6.2).

Onceestimated,densityandabundanceestimatescanbeusedtoinvestigatespatialandtemporalchangesin

animal distribution and population size. Spatial maps of animal abundance can be produced using density

surfacemodelling(Milleretal.,2013;BorchersandKidney,2014),whichcanbeusedtoassesspotentialimpacts

ofanthropogenicactivity(Scott-Haywardetal.,2013;Bucklandetal.,2015).Timeseriesofdensityorabundance

estimatescanbeusedintrendanalyses,ifsufficientdataareavailable(Thomasetal.,2004).Poweranalyses

canbeconductedtoassesstheamountofdatarequiredtodetectsignificanttemporalandspatial trends in

densityorabundanceacrosstime(Thomasetal.,2004).

7.1.3 Focal-follows

There may be a requirement to assess potential anthropogenic impacts at a finer resolution by studying

individualanimals.Focalanimalsurveyshavedifferentattributestopopulation-focussedsurveys.Conducting

individual-basedsurveysofmarinemammals,turtlesandfishrequiresfocalanimalstobedetected,assessed

andtracked.Reactionsoftrackedanimalstospecificsounds(e.g.inplaybackexperiments)orotherstimulican

thenbeassessed(e.g.Antunes,2014l;Miller,2014).Marinemammals,turtlesandfishhaveallbeentracked

usingtelemetrydevices(e.g.marinemammals:Curéetal.,2015;turtles:Kobayashietal.,2008;fish:Blocket

al.,2005),andvisualfocalfollowsofmarinemammalsareregularlyconductedfromvessels(e.g.Karniskietal.,

2014).Marinemammals can also be tracked using passive acousticmonitoringwhilst they are acoustically

detectable(e.g.Gassmannetal.,2015),andunderwatervideohasbeenusedtomonitorindividualturtlesand

fish(turtles:Smolowitzetal.,2015;fish:Mellody,2015).

7.2 Autonomousplatformtypes

7.2.1 Poweredaircraft(UAS)

Poweredaircraftcaneitherflyautonomouslyorbepilotedremotely(Figure2).Theequipmentcanbelaunched

fromdifferentplatformsusing,forexample,catapultsystemsandrecoveredbyusingahookandnetsystem

whendeployedfromaship,orbylandingonaflatsurfacewhendeployedfromlandorship.Theaircraftwill

autonomouslyfollowthedesignedtrackbasedonprioruploadedflightplans,oftenrelyingonpilotsfortake-off

andlandingonly.Agroundstationplaysakeyroleinmissionexecutionandplanning,asitallowsforcontroland

monitoringoftheaircraftandpayload,anduploadingandupdatingwaypointsandflightplansbeforeandduring

28


DATE:July2016


the flight. A typical ground station will have a telemetry link to the aircraft for monitoring and control, a

visualisationoftheaircraftandtheflightplanandtheabilitytoaltertheflightplanduringthemission.Here,we

include fixed-wing systems and Vertical Take-Off and Landing (VTOL) systems. VTOL systems are gaining

attentionfortheirflexibilityindeploymentandrecovery,whichovercomesomeofthelimitationsoffixed-wing

systems. VTOL unmanned aircraft categories includeMulti-copters/Multi-rotors, Quadcopters, Hexacopters,

Octocopters,andspecialisedFixed-wingsystemswithmulti-rotorsinstalledwhichcanbeusedforbothtake-off

and landing. In general, the system payload is dependent on the size of the aircraft, with larger aircraft

supportingheaviersensorsystemssuchassinglelensreflex(SLR)camerasand/ormulti-spectralandthermal

sensors.ThesesystemscanbevaluableduringanimalsurveysforE&Pactivitieswhichheavilyrelyonthecamera

quality,maximumaltitudeandflighttime(Koskietal.,2009a).

Figure2.ExamplesforUAS:Leftpicture:SkyprowlerUAS©KrossbladeAerospace,rightpicture:LandingofCryoWingUAS.PhotobyNORUT.

7.2.2 Motorizedgliders(UAS)

Motorizedglidersarea special classof fixed-wingaircraft capableof turningoff theengineandperforming

(parts) of the flight through soaringon rising air.Unmannedgliders are rare, andareonlyused for specific

scientificresearchinmeteorologyoratmospherescience,ormilitarypurposes(glidingbombs).Asthecolumn

ofairaboveanoceancanberegardedasasinkzone(whichistheabsenceofweakatmosphericthermalsto

pushtheaircraftupwardsormaintainaltitude),motorisedglidersareconsideredimpracticalformarineanimal

studiesandmonitoringinrelationtoE&Pactivities.

7.2.3 Kites(UAS)

Kites,whichareessentiallysmallparachutes(Figure3),consistofasteelframewithanengineandparachute

wing.Thesesystemsareeasytotransportandareabletocarrystill,videoand/ormulti-spectralandthermal

sensors. They can also operatewith auto-pilot and conduct full coverage of the surveyed areawith image

overlap. However, this equipment can only be operated in wind speeds under 6m/s, whichmay limit the

locationsinwhichtheycanbeused.Addingtothedependencyinenvironmentalconditions,thisequipmentis

often not stabilised, resulting in difficulties in hovering over an object or location of interest. This type of

platformhasbeensuccessfullyusedinEurope,AfricaandAustraliamainlyformapping,agricultureandforest

29


DATE:July2016


management(Thamm,2011).Thelimitedamountofscientificliteratureconcerningtheapplicationsofthese

systems inmonitoring during offshore operationsmay indicate the lack of tests conducted. This is possibly

associated with the weather limitations of this equipment. However, combined kites with lighter-than-air

aircraftseemtohavebecomeapopularalternative(e.g.MacKellaretal.,2013;Verhoevenetal.,2009).These

combinedsystems,socalled"kitoons"havethequalitiesofbothsystemswithfewoftheirlimitations.Though

tethered,theyareabletoendurestrongerwindswhilekeepingarelativelystablepositionduetotheirballoon

bodyandattachedsail,whichhighlightstheirpotentialformarinedataacquisition.

Figure3.KiteSUSI62take-off.PhotobyThamm,2011.

7.2.4 Lighter-than-airaircraftsystems(UAS)

Lighter-than-airaircraftsystems(blimpsorballoons,Figure4)consistofabuoyantgas-filledballoon(e.g.helium

orhydrogen)withaconnectiontosupport imagerypayloadwithstraight-downortiltedviews.Thetethered

balloonisthenconnectedandtowedbyavessel(Hodgson,2007).Themaindifferencebetweenablimpanda

balloonisthelackofatetheringsysteminblimps,whichmakesthemlessstableinharshweatherconditions

butenablesthemtoconduct independent flightswithoutrelyingonasupportvessel.Therequiredsizeofa

blimpisdependentonthepayloadweightneeded.Operatingasmallerandlighterpayloadreducesthecostsby

minimisingtheamountofheliumrequired,butmayresultinlowerimagequality.Amechanismtodetermine

distance is not yet incorporated in these pieces of equipment (Hodgson, 2007). However, the systems are

capableofcarryingGPS technology thatmay improvepositionaccuracyandestimatedistance to the target

object.

30


DATE:July2016


Figure4.Thelighter-than-aircraftsystem"blimp-cam"andcomponentsnecessaryforflights.PhotobyHodgson,2007.

7.2.5 Propellerdrivenunderwatercraft(AUV)

Propellerdrivenunderwatercraftaregenerallyrelativelysmallautomatedvehicles(seeFigure5),manoeuvrable

in the water column in three dimensions by an on-board computer. Although these systems are normally

deployedfromavessel,theyarenotphysicallylinkedtoit(untethered)andperformthedatacollectioninan

autonomousmanner(Wynnetal.,2013).

MostpropelledAUVfollowatraditionaltorpedoshape,whichoffersthebestcompromisebetweensize,usable

volume,hydrodynamicefficiencyandeaseofhandling(Vijay,2011).OthershapesfoundinpropelledAUVare

teardrop, oblate, rectangular or open space frame (typically twin-hull). They range in size from a portable

lightweightAUVtolargediametervehiclesover10mlength(Vijay,2011).Thevehicletypicallyfollowsaprecise

pre-programmed course and canoperate forperiodsof a fewhours to several days inmost environmental

conditions.Somenavy-developedmodelshavebeencontrolledviabidirectionalradiooracousticdatalinksand

somehavebeenprogrammedtomake“smart”navigationdecisionsbasedontheirownsensordata.Propelled

AUVoperateincruisemode(automated),collectingdatawhilefollowingapre-plannedrouteatspeedsbetween

1and4knots(Vijay,2011).TravellingspeedsofthepropelledAUVarecomparabletotidalcurrents,whichcan

producenavigationaldriftandtherebyaffectthedataquality.ThismakespropelledAUVlesssuitedtoshallow

wateroperations (Wynnetal.,2013).MostpropelledAUVcanoperateatdepthsofover200m,withsome

modelsabletoreach6,000mdepth.PropelledAUVkeepalineartrajectorythroughthewater,whichisessential

forseabedmappingandsub-bottomprofiling.However,foroptimalmanoeuvrability,theweightandbalance

ofagivenAUVcraftisasignificantconsideration.

31


DATE:July2016


Figure5.PropellerAUVREMUS100inafieldstudy(left,byAnreyShcherbina,WHOI)anddetailofthemodel(right,KongsbergMaritimewebsite).

Propeller based thrusters, particularly Kort nozzles, are themost commonpropulsion systemsused inAUV.

Thesethrustersareusuallydrivenbyelectricmotors,poweredbyrechargeablebatteries.Batteriescanbeused

forextending theendurance (e.g.manganesealkaline, lithium),but thisalso increases thecostpermission.

Somelargervehiclesmaybepoweredbyaluminiumbasedsemi-fuelcellsforahigherenduranceandpower

output.Silver-zincbatteriesallowtheAUVtocovera longerdistance,but lithium-ionbatteriesarethemost

cost-effective(longerlife).Othertypesofrechargeablebatteriesaresealedleadacid,nickelcadmiumandnickel

hydride(Fernandesetal.,2003).Mostoftheinternalspaceisusedbythepowersource(e.g.batteries)andthe

navigationcontrolinstrumentation.However,AUVtypicallyfollowmodulardesigntoenablecomponentstobe

changedeasilybyoperators(Vijay,2011).

7.2.6 Buoyancygliders(AUV)

Autonomous underwater buoyancy gliders (hereafter UW gliders, Figure 6) are slow moving (~0.5 knots),

typically small (<2 m), low power platforms that house sensors capable of making multidisciplinary

oceanographic observations over months and hundreds to thousands of kilometres (Davis et al., 2002).

Underwaterbuoyancyglidersmovethroughthewatercolumnbychangingtheirvolumeandbuoyancytoprofile

verticallyintheocean.Variationsinvolumeareachievedthroughinflatingordeflatingtheiroilorwaterfilled

bladder.Wingsareusedtoconvertverticalvelocityintoforwardmotion,sothattheglidertypicallyfollowssaw-

toothprofiles.Thevehicleissteeredeitherbychangesinthecentreofgravityrelativetothecentreofbuoyancy

or by using a rudder.When at the surface, satellite navigation and communication enable the glider to be

directedandcontrolledremotelyandperformadownloadofdata.

Therearecurrentlythreeclassesofunderwaterbuoyancygliders(WoodandMierzwa,2013):1)thosethatuse

mechanicalorelectricalmeansofchangingtheirbuoyancy(i.e.dropweightsorelectricalpowerfrombatteries),

2)thosethatusethethermalgradienttoharnesstheenergytochangethevehiclesbuoyancy,and3)hybrid

vehiclesthatcombinestandardpropulsionsystemsandbuoyancyglidersystems.

32


DATE:July2016


Figure6.UnderwaterbuoyancyglidersSeaglider(left)andSlocum(right)(images©DamienGuihen).

7.2.6.1 (Mechanical)Electricbuoyancygliders

Electric buoyancy gliders change their buoyancy by converting battery energy into a volume changewith a

hydraulic pump. Energy efficient movement is enabled through using only slight changes in buoyancy to

minimizeexpendedenergyandmovingslowlytoreducedrag.Fairing(theexternalsurface)shapeisdesigned

todecreasevehicledrag.

7.2.6.2 Thermalbuoyancygliders

Inanelectricbuoyancyglider,approximatelyeightypercentofvehiclepowermaybeusedforballastpumping

toalterdensity(Graver,2005).Thermalbuoyancyglidersusethedepth-relatedvariationinoceanictemperature

toeffectchangesinplatformdensity.Volumechangeinaphase-changematerialsuchaswaxisusedtoalter

buoyancyasitvariesbetweenaliquidandasoliddependingontemperature.Extrapolationofenergyuseduring

ashortdeploymentofaSlocumthermalinMarch2010indicatedapotentialenduranceof3.4years(Joneset

al.,2011a),greaterthanthreetimesthedurationoftheelectricbuoyancyglider.Initiallyjustusedforpowering

thebuoyancychange,researchanddevelopmentisnowexaminingthepotentialforharvestingthermalenergy

topowerthesensors(Jonesetal.,2011a;Buckleetal.,2013).

7.2.6.3 Hybridbuoyancygliders

Thehybridbuoyancyglidercombinesthebuoyancypropulsionmethodwithjetorpropellerpropulsion,where

internalenergy isconvertedtopropelleror jetthrust.Buoyancyglidersareconstrainedbyarequirementto

undertakeverticalsawtoothprofileswhenmovingforward.Jet/propellerpropulsionenablesgliderstomake

33


DATE:July2016


constant altitude transits and can improve their ability tomake progress against strong ocean currents, or

navigateareaswithrestrictedaccesstosurfacewaters(Claus,2009).

7.2.7 Poweredsurfacecrafts(ASV)

Unmannedpoweredsurfacecraftoffergreatversatilityandanumberarenowavailable.Thereisadiversityof

approachesanddesignsandtheytakemanyforms.Mostoftenthesevehiclesareoperatedremotelyfromshore

orasupportvessel.Suchcraftmayalsobeknownas“autonomous”.Asadefiningterm,“unmanned”ismore

technicallycorrectasnoneofthesevehiclesareautonomous inthesenseofbeingwholly independent–all

require some mission planning and/or remote operation by human hand. The following examples of

autonomous powered surface craft are categorised into three groups based solely on hull shape: boats,

catamaransandsemi-submersibles.

7.2.7.1 Boat

ManysuchASVbearastrongresemblancetoanotherwisefamiliarcraft,suchasarigid-hulledinflatableboat

(RHIB)withaninboardoroutboardmotor.Essentially,anysmallcraftcanbepotentiallyconvertedintoanASV

bytheadditionofcontrols,navigationandtelemetry.Forlegalreasons,duetothepotentialfornavigationalrisk

toothercraft,boatscannotbefullyautomated,andhavetoincorporatethepossibilityofmannedoperation.

Oneofthemostpopularwaystocreateautomatedordual-mode(manned/unmanned)vesselsistointegrate

manual-to-automatedcontrolconversionkits intocommercialvessels(Caccia,2006;Cacciaetal.,2009).The

advantageofthisapproachisthatremotecontrolbringsarangeofpossibilitiestofamiliarvesselsofgreatersize

large and complexity. The disadvantage is that the increased complexity may bring issues on mechanical

reliability,whichcannotberesolvedremotely.

Figure7.AnAUVboattypeViknesdesignedanddevelopedbytheNorwegianUniversityofScienceandTechnologyinconjunctionwithMarineRobotics(image©NorwegianUniversityofScienceandTechnology).

34


DATE:July2016


7.2.7.2 Catamarans

Forautonomousoperations,acatamaranbasedhullgivestheabilitytocarrysignificantpayloadonalargerarea

of deck space. This enables ready access to the payload and a flexible modular build approach as well as

supportinglargetop-heavyunits,suchasatelemetrymast.Acatamarangivesaverystableat-seaplatformbut

thismaybeat theexpenseofmanoeuvrability– themaindisadvantagecitedof thishull-type.Acatamaran

shapeisreadilyscalableinsizehoweverandthisallowsforscopeinambitionandcost.Smallermodels,suchas

theSpringer(Figure8left),areidealforresearchbuthavealsobeenutilisedforcivilapplicationsandeducation

purposes.ThelargerASVC-Enduro(Figure8right)utilisesitsshapeanddeckspacetocarrysolarpanelsand

supportathree-bladewindturbine.Missionrangeistherebyincreasedandspacestillremainsfortheaddition

ofsensorpackages.Thishas includedSeichePAMwhichunderwenttrials in theGulfofMexico in2015and

successfullydetectedanimalsremotelyandinreal-time.

Figure8.Catamarans:Leftpicture:Springer,acatamaranbasedASVdevelopedbyPlymouthUniversity(image©PlymouthUniversity),rightpicture:ASVC-Enduro,alargescalecatamaranbasedASV(image©ASV).

7.2.7.3 Semi-submersible

Themainbodyofthesemi-submersiblevehicle(Figure9)remainsunderwater.Onlyasmallsurfaceexpression

isgivenwiththecommunicationsmastandairintakeprotrudingabovethewaterline(Manley,2008).Thecraft

isnecessarilybulkyandtheeffectistoreducetheeffectsofwavesandmakeitverytolerantinroughseas.The

vehicleoffersaparticularlyhighlevelofsea-goingcapabilitygivenitsrelativesmallsizeandcost.Italsoenables

excellent seakeepingcapabilities,beingable to remainat itsdesignwaterline inalloperational speedsand

conditions,withoutrelianceonhydrofoilcontrol.Thesemi-submersible’sbulkalsolendsithighsurvivability.On

thedownside,physicallymanagingthevehicleduringlaunchandretrievalremainspotentiallychallenging.The

design is primarily intended for application in naval use but has potential for conducting bathymetric and

hydrographicsurveysincivilandcommercialsectors(Wolking,2011).

35


DATE:July2016


Figure9.Thesemi-submersibleASV-6300designedbyC&CTechnologiesandASV(imagefromWolking,2011).

7.2.8 Self-poweredsurfacevehicles(ASV)

Self-poweredsurfacevehiclessuchaswaveandwindglidersuserenewableenergyfromwave,windorsolar

energy for generating forward motion or station keeping. The Liquid Robotics Waveglider (Figure 10), for

example,isnowawell-establishedautonomoussurfacevehicle.Itssurfacefloatcontainssolarpanelstopower

thenavigationsystemsandsciencepayloads,whileforwardpropulsionisprovidedbyarackoffinssuspended

7metres below the floatwhich directly convertwavemotion into forward power. The latestmodel of the

Waveglider,theSV3,alsohasasmallelectricallydrivenpropeller.Thecombinationofwaveandsolarpower

meansthatthesevehiclescanstayatseaindefinitelyanddeploymentsofmanymonthshavebeenreported.

Otherwaveandwindpoweredvehiclesarealsobecomingcommerciallyavailablesuchasthewindpowered

SailBuoy1andthewavepoweredAutonaut

2and is likelythatothervehicleswillappearofthemarket inthe

comingyears.

1 http://www.sailbuoy.no/ 2 http://www.autonautusv.com/

36


DATE:July2016


Figure10.UnderwaterphotographofaLiquidRoboticsWaveglidershowingthesubunit,whosemoveablefinsconvertverticalmotionfromwavesintoforwardpowerandthesurfacefloat(photo,LiquidRobotics).

7.2.9 Drifter(ASVandAUV)

Driftingsensorpackageshavelongbeenusedbyoceanographers,oftenbecausetheirlowcostmakesitpossible

todeploymanysensorsatonce,therebyincreasingsamplesize.Manydrifterssimplystayatthesurfaceand

relay theirpositionsviasatellite link inorder tostudyoceancurrents.Moresophisticateddevicesalsocarry

sensorpackages,andinthecaseoftheARGOfloats3,moveupanddownthroughthewatercolumn,sampling

atvaryingdepths,beforetransmittingtheirdatatoshore.

TheclassicexampleofadriftingPAMsystemisthesonobuoy.Thesehavebeenusedbythemilitarysincethe

1940s,primarilyforsubmarinedetection.Modernsonobuoysgenerallyconsistofacylindricalpackagethrown

fromanaircraft.Whentheyhitthewater,ahydrophonedeploystoasetdepthandaVHFtransmitterisused

toradiosoundsbacktothecirclingaircraftwhichmustremainwithinlineofsightofthebuoy.Constellationsof

buoyscanbeusedtolocalizesoundsources,thoughlifetimesaretypicallyonlyafewhours.

Mostautonomousdriftersystemsarenon-recoverable,withallimportantdatabeingtelemeteredtoshoreor

satellitelink,whichpresentsachallengeforacousticmonitoringdataduetothehighvolumesofdatainvolved.

Griffiths (2015) describes a low-cost drifting sensor package for passive acoustic monitoring which can be

constructedforaround$5,000.Thissystemcontainsatwo-channelrecorderanddataarestoredonboardthe

device,onlyavailableonceithasbeenrecovered.

3 http://www.argo.ucsd.edu/

37


DATE:July2016


7.3 Sensortypes

7.3.1 Electro-opticalimagingsensors

There are fourmain imaging technologies that should be evaluatedwhen considering detection ofmarine

animalsfromaircraft;RedGreenBlue(RGB),thermalInfra-Red(IR),non-thermalIR,andvideocameras.These

foursensortypesdiffermainlyinwhatpartoftheelectromagneticspectrumtheycanregisterradiationfrom.

ThoughRGBcamerasareabletocapture infraredwavelengths, for thepurposeof thisevaluationweassign

infraredcamerastoitsspecificcategory,includingboththermalandnon-thermalinfrared,giventhatthereare

sensorsdevelopedforthesinglepurposeofcapturingIRwavelengths.Severalsensorsarecurrentlyavailable

forUAS.Dependingonthepayloadavailable,asingleUASmaycarryoneorseveralsensorstoincreaseaccuracy

ofdetections,whichcanprovide informationonobjectpresence/absenceandanimalbehaviour/movement.

The informationgatheredbythesesensorscanbetransmittedtothegroundinnearreal-timeorstoredon-

boardtheaircraftforpost-processing.Byutilisingthedatawithregardtopositionandorientationofthecamera,

foreachimageorframe,itispossibletogeoreferenceeachindividualpixelinanimageinrelationtoageographic

coordinatesystem.Thisissometimesalsoreferredtoasgeocoding.Fordetectinganimalsfromanunmanned

aircraftitisimportanttomaximizethefieldofviewwhilemaintainingsufficientsurfaceresolutionandcontrast.

Todetectsmallmarinemammalssuchasporpoiseswhilesurfacingasurfaceresolutionofapproximately20

cm/pixelisrequired.

7.3.1.1 RGBcameras

AnRGBcameraconsistsofmillionsofsmallelectro-optical(EO)sensorelements,referredtoaspixels,which

register incoming radiation focused througha lens, and convert it to anelectric charge. This charge is then

digitisedandstoredonthecameraortransmittedtoacomputer.Camerascandetectradiationindifferentparts

ofthespectrumbutmostcommercialcamerasonlydetectlightinthevisiblerange(VIS)oftheelectromagnetic

spectrumbetween400to700nm.Thetwomostcommoncamerasensorstodayarethecharge-coupleddevice

(CCD)andthecomplementarymetal-oxidesemiconductor(CMOS).Tobeabletoseparatecoloursinanimage

usingasinglecamerasensor,theindividualsensorelementsarefittedwithared,greenorbluecolourfilter

array.Cameraswithnofilterarereferredtoasmonochromecameras.Theycannotseparatecoloursbutbenefit

from a much higher sensitivity than a colour camera. Commercial cameras come in a variety of different

qualities,withregardtoresolution,dynamicrangeandsensitivity.Camerascanrecorddatawithahighframe

rate,which,combinedwithahighresolution,generateslargeamountsofdata.

7.3.1.2 ThermalIRcamera

Thermal camerasor sensors register thermal radiationemitted fromanobject. Thermal imagers aremainly

dividedintwocategoriesbasedonthewavelengthregiontheyoperatein(1)mid-wavelengthinfrared(MWIR)

imagersoperatingintherangeof3to5µmand(2)long-wavelengthinfrared(LWIR)imagersoperatinginthe

38


DATE:July2016


rangeof8to15µm.LWIRimagersaremostcommoninsmallandmediumsizeUASduetotheirsmallsizeand

relativelowcost.However,theresolutionislimitedandatypicalhighendLWIRimagerhasaresolutionof640

x512pixels(Tau640,FLIRSystems).Theenergyemittedfromanobjectwithregardtofrequencyorwavelength

canbeestimatedusingthePlanck’slawforblackbodyradiation(Planck,1914),wheretheenergyisshiftedto

longerwavelengthforlowertemperatures.Hence,forobjectssuchasmarineanimals,LWIRimagerswouldbe

theappropriatesolution.

Water isnon-transparent to thermal radiation, anda thermal IR systemcannot seeanythingbelow the sea

surface (evena fewmicrometres).Hence, theuseof IR systems formarineanimalobservationandstudy is

limited to surfacing animals only, resulting in the exclusion of fish and turtles as study objects (except for

leatherbackturtles,whichseemtobeanexceptionastheyregulatetheirbodytemperaturesimilartomammals

(Bostrometal.,2010)).Athinlayerofwateroftencoverssurfacinganimalsandexperimentshaveshownthat

when a surfacingminkewhale (Balaenoptera acutorostrata) is covered by a thin film ofwater, its thermal

radiationisalmostcompletelymasked;thetemperaturedifferencebetweentheanimalandthesurrounding

waterisusuallylessthan0.1°C(Baldaccietal.,2005).However,theblowandblowholeareeasiertodetectand

temperaturedifferencesofover4°Cbetweenblowholeandsurroundingwaterhavebeenreported(Baldacciet

al., 2005). Several publications have investigated the use of IR cameras for detecting blows using thermal

imagers;blowsfromlargebaleenwhalesweresuccessfullydetectedupto7kmaway(SanthaseelanandAsari,

2015; Weissenberger et al., 2011; Zitterbart et al., 2013). These studies all applied a higher quality (third

generationorlater)thermalimagerplacedonalargeshiporland-basedinstallation,allowingaviewoftheblow

againstthewaterbackground.Additionally,studiesfrommannedaerialsurveysprovideevidencethatitisalso

possible to detect thermal "footprints" of animals surfacing, using the same typeof equipment (Churnside,

Ostrovsky,andVeenstra,2009).

UsingthermalIRfromUASfordetectionofmarineanimalsseemsplausibletodetectmarinemammalsboth,

duetothelargethermaldifferencebetweenthecetaceanblowholeandthesurroundingsea(seealsoVerfuss

etal.,2016),anddifferencesinbodytemperatureofsmalleranimals.Inpracticehowever,wehavenotfound

any publication documenting the use of thermal imagers onUAS formarinemammal detection. Awhale’s

blowholeiscomparativelysmallandwouldrequirearesolutionofacoupleofcm.Forachievingaresolutionof

2cm/pixel,theresultingwidthofthescenecoveredbyathermalimagerwith640pixelswouldonlybe12.8m.

Toachieveanappropriateresolution,severalcamerascouldbecombined,whichwouldresultinabulkysolution

notwellsuitedforsmallormediumsizeUAS.ThermalIRisapromisingtechnology,thoughthecombinationof

payloadrequirementsofUASandcameraresolutionforaerialdetectionthatiswithinreasonablepricerange

forunmannedsurveys,ispossiblythemainreasonforwhythishasyettobedevelopedfurther.

7.3.1.3 Non-thermalIRcamera

Non-thermalIRcamerasoperateinthelowendoftheIRspectrumandareoftendividedintotwogroups:near-

infrared(NIR)intherangeof0.74to1.0µmandshort-wavelengthinfrared(SWIR)intherangeof1to3µm

39


DATE:July2016


(Maldague,2007). Inthisregionofthespectrumverylittleenergyisemittedfromobjectswithnormalbody

temperatureandthesecamerasworksimilarlytoRGBcameras,whicharebasedon lightreflectedfromthe

object.SWIRcamerasexcelinlowlightconditions,partlybecausethesesensorsarehighlysensitive,butalso

duetoaphenomenonreferredtoasnightskyradiance.Thenightskyemitsfivetoseventimesmoreillumination

thanstarlight,nearlyallofitintheSWIRregion.Hence,withaSWIRcameraonecan“see”objectswithgreat

clarity onmoonlessnights. SWIR cameras are available in small sizes, such as the TAUSWIR (FLIR Systems)

weighting only 130 grams.However, SWIR cameras share similar limitations as thermal LWIR cameraswith

regardstorelativelowresolution(nomorethan640pixels).WehavenotfoundanyworkwhereSWIRcameras

havebeentestedforthedetectionofmarinemammals.

7.3.1.4 Video

TheuseofaHDvideowitharesolutionof1,920x1,080pixelsatanoperationalaltitudeof122m,andaground

resolutionof20cm/pixelcrossresultsinacross-trackdistanceof384mtrack(stripwidth),basedonoperations

within the Visual Line of Sight (VLOS). Flights at higher altitudes will cover larger areas, though national

authoritiesrequirespecialpermitstooperateatlargerrangesandaltitudes.Oneofthemostimportantfeatures

forthedetectionofmarineanimalsfromUASusing livevideo isthatthecamera isstabilisedusingagimbal

(Koskietal.,2009a).Thiswillreduceimagevibrationandimprovetheefficiencyoftheoperatormanuallytrying

toidentifyanimalsinthevideo.Reducingimagevibrationwillalsoreducethevideobitrateduetoimproved

encodingefficiency.Thoughitisalsopossibletostorethevideocollectedforfutureanalysis,thetransmission

ofreal-timedatatothegroundstationcanprovidevaluableinformationthatcanbeusedforbothscienceand

management(e.g.Hodgsonetal.,2013;Koskietal.,2013;Koskietal.,2009a).

7.3.1.5 Otherelectro-opticalimagingsensors

SensorscollectingLIDAR(Light-DetectionandRanging)remotesensingdatausepulsedlaserlighttomeasure

rangestotheEarthbyusingultraviolet,visibleornearinfraredlighttoimageobjects(NOAA,2015).Thistypeof

sensorhasprovenitsvalueinsurveysofmannedaircraftanditisgaininginterestintheUASmarket.LIDARhas

shownitsrelevanceinfisheriesandcandetectfishatdepthsupto16meters(Butler,1988)thoughthedetection

of marine organisms using such technology remains untested for UAS surveys, possibly due to the same

limitationsasforinfraredequipment.Despitetheinterestintechnologiesusinggreenlightthatcanpenetrate

thewater surface, thisequipmentuse currently appears tomainly focuson topography (including sea floor

elevations)andhydrodynamicmodelling.Thus,theuseofthistypeofequipmentisnotfurtherdealtwithinthis

report.

7.3.2 PAMsystems

Passiveacousticmonitoringreliesondetectingthesoundsproducedbyanimals.Threegroupsofmarineanimals

producesound:crustaceans(predominantlyshrimp),fishandmarinemammals(Horne,2000).Snappingshrimp

40


DATE:July2016


aredominantsoundproducersinshallowwatersatlatitudesoflessthan40°(Horne,2000),asinglesnapcan

reachpeak-to-peaksourcelevelsof185dB(re1µPa)withabroadfrequencyspectrumupto200kHz(Auand

Banks,1998).Soniferous(soundproducing)fisharecommon,andcanbeusedtoidentifyspecificspecies.Over

800speciesoffishfrom109familiesworldwidearecurrentlyknowntobesoniferous(Kaatz,2002).Fishproduce

soundstocommunicatewitheachotherwhilefeeding,matingoraggressivedisplaysaswellasmakingincidental

noisesassociatedwithfeeding,swimmingandotherbehaviours(Rountreeetal.,2006).Thesoundsoffishesare

rarelyhigherthan1kHz,atleastinthezonesofmaximumamplitude,andincludethoseproducedbyfriction

between bones, contractions of sonic muscles, or hydrodynamic noises from swimming. A comprehensive

reviewofsoundandsoundproductioninfishescanbefoundinKasumyan(2008).Itisgenerallyconsideredthat

PAMcouldbeusedtomonitorthedistribution,abundanceandbehaviouroffish(Rountreeetal.,2006).The

useofPAMfor thedetectionof fish isstillverymuchadeveloping fieldofactivity,andwhile theremaybe

potential for monitoring some species and spawning events, active acoustic survey and tagging fish with

ultrasonicsoundemittingtagsremainthemorecommonmethodsusedtostudyfishspecies.

Thespeciesgroupforwhichpassiveacousticmonitoringismostwelldevelopedismarinemammals,particularly

cetaceans (whales anddolphins).All cetacean species are known tomake soundsof some typeor another,

althoughforseveralspecies,soundproductionislimitedtocertainindividualsatcertaintimesoftheyear(for

instance,malehumpbackwhalesingingduringthebreedingseason).Forsomespecies,particularlydeepdiving

odontocetes,individualsspendsignificantlymoretimevocallyactiveandavailableforacousticdetectionthan

theyspendatthesurfacewheretheywouldbeavailableforvisualdetection.Ontheotherhand,manyspecies,

orimportantcomponentsofpopulations,areknowntoremainsilentforextendedperiods,thusmakingthem

unavailableforacousticdetection.

The frequency rangeofmarinemammals vocalisations span14octaves from the infrasonic soundsof large

baleenwhales,witheachsoundlastingformanyseconds,toshort(100µs)ultrasonicecholocationsignalsfrom

smallodontocetesatfrequenciesrangingupto150kHz.Forexample,individualvocalisationsofabluewhale

maylast150,000timeslongerthanthoseofaharbourporpoise,andthefrequencyofaharbourporpoiseclick

maybe15,000 times thatof abluewhale call. This incrediblywide rangeof sound types createsparticular

problemsforsystemsattemptingtodetectandlocalisemultiplespeciestypes.

VocalisinganimalscanbedetectedsolongasthePAMsystemhassufficientbandwidthtocapturethesounds

of interest. Sampling theory dictates that the sample ratemust be at least twice the highest frequency of

interest,soforsmallodontocetes,suchastheharbourporpoise,sampleratesintheseveralhundredsofkHz

arerequired,witha500kHzsampleratebeingtypicalformanysystems.

Detectionrangeisgovernedbyhowloudtheinitialsoundis,howmuchnoisethereisaroundthereceiverinthe

frequencybandofinterestandtransmissioneffectssuchasdownwardrefractionandhigherabsorptionofhigh

frequencysounds.For largebaleenwhales,detectionranges inthetensorevenhundredsofkilometresare

possibleandspermwhalesareoftentrackedatrangesofseveralkilometres.Smallerspeciesthatproducehigher

41


DATE:July2016


frequencysignalstendtohavemuchshorterdetectionranges,ofteninthelowhundredsofmetreseveninlow

noise conditions. Detection range can be expected to vary dramatically depending on local industrial noise

sources,suchasairgunsandvesselpropulsionsystems.Certainlyabigpotentialadvantageofsmallautonomous

vehicles for PAM is that the local noise field can be expected to be both low and consistent between

deployments,meaning thatdetectionperformancemaybebothbetterandbetterunderstood than it is for

manyvesselbaseddeployments.

Acousticsourcescanbelocalisedusinganumberoftechniques.Bearingstosoundscanbeobtainedwithsmall

clustersofhydrophonesand forsomespecies,multiplebearings to individualanimals, taken fromamoving

platform,canbeusedtoestimaterangeasisdemonstratedforspermwhalesinLewisetal.(2007).Asageneral

rule,thelowerthefrequencyofthesoundofinterest,thewidertheseparationbetweenhydrophonesneeded.

Separationsoftensofcentimetresareadequatetoobtainbearingstobroadbandorhighfrequencyodontocete

clicks,butseparationsofmanymetreswouldbeneededtoestimatebearingstolowfrequencybaleenwhale

calls.Inordertomeasurerangedirectlytothesourceofasinglesound,widelyspacedhydrophonesarerequired

andlocalisationwillonlybeaccuratewithinapproximately3timesthearraydimension(i.e.ifhydrophoneswere

spreadover100m,localisationwouldbeaccuratetoabout300m).Separatinghydrophonesbysuchdistances

canpresenttechnicalchallengesonboardavesselandwouldbeparticularlyproblematiconsmalllowpowered

autonomousvehiclesystems.

MostPAMsystemsincludeahumanoperatorinanydecision-makingprocess,whetherthatbereal-timedecision

makingorpostanalysisofdata.Bytheirverynature,autonomousplatformsdonotaccommodateahuman

operator,sodatamustbestoredortransmittedinaformwherebyitremainspossibleforanoperatortoview

sufficientdatafordetectionverificationpurposes.

PAMsystemsforautonomousdatacollectioncanbroadlybedividedintotwocategories:

1. Rawdatasystems

2. Systemswithonboarddataprocessing.

Thelargesizeofmodernharddrives,andmorerecentlyofflashmemorycards,hasledtothedevelopmentof

awidevarietyofsystemswhichsimplystreamrawdatatostoragemediaforlateranalysis.Fixedautonomous

recordersforPAMwerethoroughlyreviewedinanindustry-fundedprogramin2013(Sousa-Limaetal.,2013)

anditisnotourintentiontorepeatthisreviewprocesshere.Asisclearfromthiswork,thereisageneraltrade-

off between size and capability in these systems.Many of these systems are nowwell established and are

commerciallyavailableforhireorpurchase.Whileintendedforfixeduse,manycouldpotentiallybeadapted

(e.g.throughtheadditionofanexternaltowedhydrophone)foruseonmovingautonomousplatforms.

Packageswithonboardprocessingare lesscommon,althoughtherecentavailabilityofpowerful lowpower

usageprocessors,oftendevelopedforthesmartphonemarket,meansthatthisisarapidlydevelopingfield.

Whenagliderisatthesurface,summaryinformationofcandidatedetectionsissenttoshoreviasatellitelink

42


DATE:July2016


whereitcanbeverifiedbyahumanoperatorfornearreal-timedetection.ThesystemdevelopedbyKlinkto

detect beaked whales also sent back some summary information about detections during each dive, but

insufficientdataweresenttoshorefordatavalidation.Datavalidationwasthereforeonlyconductedoncethe

vehicle was recovered and full recordings analysed. Similarly, Gillespie deployed a Decimus system on a

Wavegliderforthedetectionofharbourporpoisesandalthoughsummarycountsofdetectionsweresentto

shoreviasatellite,detaileddataneededtoconfirmthosedetectionswereonlyavailableoncethevehiclehad

beenrecovered.

As devices increase in acoustic bandwidth, so the pressures on data storage, data processing and power

requirementsbecomemoreacute.Thiscanlimitthepracticalitiesofusingthesesystemsforlongmissionson

small,lowpowervehicles.

7.3.3 AAMsensors

Anactiveacousticsensoractsbybroadcastingasoundwaveinthewaterandmeasuringthereturnedsignal

fromencounteredtargets.Theperformanceofthesonarsystemdependsonthedegreetowhichthebeamof

soundisfocussedontothetargetandhencethedirectivityofthearraygeneratingthesoundbeam.Itisalso

dependentonthelevelofbackgroundnoise.Activeacousticmonitoringequipmenttypicallyincludesfisheries

sonarandechosoundersthatoperatefrom~18kHzthroughto~500kHz.TargetsthatcanbedetectedwithAAM

thatareofinterestforthisreviewaremarinemammals,turtlesorfish.Theanimalisdetectedthroughtarget

reflectionratherthanvocalisation.Therefore,activeacousticmethodsarenotlimitedbypoorlight,visibilityor

mostweatherconditionsandanimaldetection isnotdependentonvocalisationorsurfacingbehaviour.The

rangeanddetectionabilitiesofactivesonarisfrequencydependent;attenuationofsoundamplitudeincreases

withincreasingfrequencyleadingtoshorterdetectionranges.Ontheotherhand,theresolutionincreasesas

frequencyincreases,thereforehigherfrequenciescandetectsmallertargetscomparedtolowerfrequencies.

InordertouseAAMtodetectmarinemammals,turtlesorfishtwofactorsarerequired:amethodofidentifying

theorganismintheacousticdataandanestimateofthesensitivityoftheinstrumentandeffectsofthephysical

environmentonsoundpropagation(seeVerfussetal.,2016forfurtherdetails).Identifyingtheorganisminthe

acoustic data can be done through either target strength (e.g. Bernasconi et al., 2013), target aggregation

features(Fernandes,2009),targetmovement(stationaryfishschoolsversusmovingwhales;(Selivanovskyand

Ezersky,1996;Bernasconietal.,2013)ormulti-frequencyresponses(KorneliussenandOna,2003).

Therearefourtypesofechosounders/sonarsusedinactiveacousticresearch:single-frequencyechosounders

(allowing for multiple transducers e.g. multi-frequency), multibeam echosounders, wideband/broadband

echosounders and sonars. A generic feature of all these echosounders/sonars is that the amount of echo

returnedfromatargetisafunctionofthefrequencyusedandthesize,shape,compositionandbehaviourof

thetarget(Horne,2000).

43


DATE:July2016


7.3.3.1 Single-frequency/multifrequencyechosounder

Thesearethestandardfisheriesechosounders,typicallyavailablefrom18kHzto>500kHz,andnormallyhave

afocussednarrowbeamwidth.Thesizeofthetransducerisrelatedtothebeamwidth,withasmallerbeamwidth

resultingfromalargertransducerfaceatanyparticularfrequency(MacLennanandSimmonds,1992).Thefirst

echosoundersonlytransmittedasinglebeamofsound,wherethepositionofthetargetwithinthebeamwas

unknown.Modernechosoundersaresplit-beam,wherethetransducerhasfourquadrants,allowingthelocation

oftargetsinthreedimensions.Theuseofsingle-frequencyacousticsislimited,asitcannotdistinguishbetween

sizeandtargettypeandthereforehasreducedcapabilityforspeciesidentification.Multifrequencyacousticsare

required for species identification (KorneliussenandOna,2003), exceptwhere targets are simpleorexhibit

particularrepetitiveandrecognisablepatterns.

7.3.3.2 Multibeamechosounder

Amultibeamechosounderemitssoundwavesinafanshape.Thisgreatlyenhancestheareaandvolumeofthe

acousticwindow,typicallyprovidingahorizontalswathofinformationwithangularcoverageto150°,formed

fromhundredsofbeams.Theyimageasynopticsliceofthewatercolumnandcanensonifywholeaggregations

offish,aswellasdiscriminatingandtrackingthepositionofindividualtargetswithintheswath(Mayer,2006;

Colboetal.,2014).Mostrecently,dedicatedbiologicalsystems,whichcollectcalibratedacousticbackscatter

datathroughoutawholefanof500beamshavebeendeveloped(e.g.theMS70,Korneliussenetal.,2009).For

the detection ofmarinemammals and sharks,multibeam echosounders (vertically downward looking) and

multibeamimagingsonars(horizontallylooking)havealmostexclusivelybeenused.

7.3.3.3 Wideband/broadbandechosounder

Widebandechosoundersmakeuseofachirppulsetocoverawidefrequencybandinasingleping,thereby

providing echo strength measurements over a wide frequency range. Wideband equipment4 currently

represents the forefront of technical developments in fisheries acoustics. It has four advantages over

narrowbandsystems:(1)spectralinformationforspeciesidentification,(2)improvedtargetdetection,(3)more

stableestimateofsignaland(4)improvedtargetresolution.

7.3.3.4 Omnidirectionalsonar

Anomnidirectionalsonaremitssoundwavesinalldirectionsinahorizontalplane.Thisgreatlyenhancesthe

areaoftheacousticwindow,andallowsdetectionofmarinemammalsandfishfromdirectionsnotimmediately

undertheship.

4 www.simrad.com/ek80

44


DATE:July2016


7.3.4 Animal-bornetags

Identityandmovementsof individualorganisms(marinemammals, turtlesandfish)canbemonitoredusing

transpondingacoustictags(e.g.,Voegelietal.,2001).Aminiatureelectronicpingercanbeplacedonananimal

andfollowedusinghydrophones(e.g.Wroblewskietal.,1994;Wroblewskietal.,2000).Tagstransmitatspecific

frequencies to provide a unique marker for each individual. Detections are generally made with bespoke

receivingequipmentfromthetagmanufacturer.Tagdetectionrangeandtransmissiondurationareproportional

totagsize.Obviously,akeyrequirementisthattheanimalistaggedinthefirstplace.

8 Evaluation of autonomous systems 8.1 Criteriaandmetrics

Thereareseveralexamplesinpublishedliteratureofcompleteautonomoussystems(seesection9.2),which

combineaplatform(orvehicle)withasensoranddatarelaysysteminordertoachieveaspecifictask.Forthe

evaluationwehaveseparatedplatforms,sensorsanddatarelaysystemstoenableacomparativeanalysisofthe

variousparts.However,themainemphasisoftheanalysisisassessingthesuitabilityoftheplatforms,sensors

and data relay systems (and any of their combinations) for mitigation and/or population monitoring or

monitoringofindividualanimals.

Manyoftheplatformsreviewedatleastclaimtobehighlyflexibleinhowtheyarefittedoutwithsensorsand

data relay systems, and data relay systems in particular must often be chosen for particular applications.

Therefore,aswellasdefiningwhatisrequiredintermsofperformance,wehavegatheredpracticalinformation

to inform which platform-sensor-transmission system combinations may be feasible in the future. General

informationontheplatformsandsensorswascollectedaswellasdetailssuchastechnicaldataandoperating

figures,expenses,mainpropertiesdeterminingmitigationmonitoringandmarineanimalsurveycapabilities,

operational information and data sources. To collect this information in a comparable manner, evaluation

criteriawerechosen.Thischapterprovidesexplanationsonhowthesecriteriaweredefinedandwhytheywere

chosen.Togenerate thecomparisonmatrices in section11.3, thesecriteriaweregrouped into the following

categories:general information,technicaldetails,costs,surveycapabilities,operationalaspectsandmanning

requirementsforboththeplatformsandsensors.Thesegeneralcategoriesfacilitatedtheorganisationofthe

hugeamountof informationgathered foreachsystem.Section11.1givesabulletpoint listoverviewof the

criteriadefinedinthischapteraswellasalinktothecorrespondingcomparisonmatrixtable.

Howaplatformandsensorareconfiguredanduseddependsverymuchonthepurposeofthecollecteddata.

Platform,sensoranddatarelayrequirementsformitigationmonitoringaswellasoperationalaspectsarequite

differenttothoseforpopulationmonitoringandthesedifferencesarehighlightedinthesectionsbelow.When

mitigating,thepriorityisgenerallyonthedeliveryofnearreal-timedataandhighdetectionefficiency,whereas

45


DATE:July2016


forlongtermmonitoringreal-timedataareoftennotrequiredanddetectionefficiencyneednotbehigh,so

longasitisreasonablyquantified.

8.1.1 Collectionofsurveydata

Toachievethesurveydesign,datacollectionandanalysisrequirementsoutlinedinsection11.6,thefollowing

evaluationcriteriawereidentified,whicharediscussedbelow.

Themajorityoftheplatformcriteriaarelinkedtosurveydesign.Aswithmitigationmonitoring,platformmission

durationandanyfactorslimitingit,andspeedarekeycriteriathatmustbeassessedtoensurethataplanned

surveycanbecompletedinthedesiredtimeframe.Theminimumandmaximumverticalrangeaplatformcan

cover(heightabovesealevelforaerialvehiclesanddepthforunderwatervehicles)arealsoimportant.Foreach

consideredplatform,itisimportanttoknowwhetherasystemhastracksettingoptionstofollowapre-designed

track,andwhetherthetrack is implementedusingpre-programmedcoordinates, throughmanualpiloting,a

combinationofbothorsomeothermethod.Environmentallimitationsthatwouldpotentiallyaffectasystem

keeping to its track are also investigated. Some systems have the ability to self-correct if deviation from a

plannedtrackoccurs.Itisimportanttoconsiderwhetherasystemcanself-correctfortworeasons.Firstly,self-

correctionwillimprovetheabilityofasystemtoadheretoaplannedsurveydesignbut,secondly,self-correction

may increase systemnoise,whichmay negatively impact data collection. The ability of a system to remain

stationary(oratleastperformturnstostayinthesamelocation)isalsoconsidered.Station-keeping(asthis

behaviourisknown)isusefulfor(1)makingvehiclesactasfixedplatformsor(2)havingvehiclesinteractwith

other fixed platforms over a small spatial scale or (3) potentially collecting fine-scale animal behavioural

information.Thecapabilityofagivensystemtomakeautonomousdecisionsand,therefore,potentiallyhave

multiplesurveymodesisalsoinvestigated.Theabilitytoswitchbetweensurveymodeswouldallowasurveyto

be adapted either through interactions with other vehicles or due to some detections/measurements

encounteredduringthesurvey.Finally,whetherornotclocksynchronisationwithothersystemsispossibleis

also a useful consideration. Clock synchronisation may be required if the same detections needed to be

identifiedacrossmultipleplatforms(thisisalsorelevanttospecificsensors).

Unlikemitigationmonitoring,itisnotnecessarythattheprobabilityofdetectionofagivenanimalorgroupof

animalsishighduringsurveys.Instead,criteriarelatingtosurveyingcapabilitiesareprimarilyconcernedwith

thepotentialtoestimatetheprobabilityofdetection,usinganyofthemethodsdiscussedinsection11.6.2.The

firstcriterionforthesensors,however,providesageneralassessmentofwhatenvironmentalfactorswould

preventoptimalsensorperformance.Thetypeofdatacollectedandstoredbythesensorisalsoascertained.It

isimportanttoknowwhetherthesensorstoresrawdataorwhetherthereissomeon-boardprocessing.Ifonly

processeddataarestored,thenitismoredifficulttoassess,forexample,theproportionoffalsedetections.It

isalsocriticaltoknowwhethertheresolutionofthestoreddataissufficienttoidentifytheobservedspecies.In

order to assess the potential of a given sensor to store data that could be used to estimate detection

probabilities,itisnecessarytoascertainwhetherthedatacouldbeusedtoestimate(1)bearingstoobservations

46


DATE:July2016


or(2)directrangeswhereabsolutelocationoftheanimalmaynotbeknownor(3)horizontalrangestoanimals

(wheresomebearingambiguitymayexist).InordertoestimatehorizontalrangeusingAUVsystems,itwould

benecessarytoknowanimaldepth.ForestimatinghorizontalrangefromUASsystems,itwouldbenecessary

tobeabletoestimatethehorizontaldistancefromthevehicle’stracklinetotheanimal.Alsolinkedtodetection

probabilityestimation,butonlyrelevantforPAMdata,arethefollowingcriteria:can(1)receivedlevelsofthe

detectedanimalsounds,and(2)ambientnoiselevelsbeestimatedfromthedata.Intheabsenceofanyother

information about acoustic detections (i.e. no range or bearing estimates), such information becomes

increasinglyuseful.Receivedlevelsandambientnoiselevels(inthefrequencybandofinterest)canbeusedto

calculatesignaltonoiseratios.Inaddition,CTDdatawillallowaccuratesoundspeedprofilestobeestimated,

whichcanbeusedinsoundpropagationmodellingtoestimatetransmissionloss.Transmissionlossisanother

acoustic measure that can be used in non-standard detection probability estimation methods. Clock

synchronisation isalsoarelevantcriteriontoassess;clocksynchronisationwillbekey if thesamedetection

neededtobeidentifiedacrossmultiplesensors.Theaccuracyofthesynchronisationwillalsodeterminewhether

multiplesystemsorsensorscanbeusedtocreateinstrumentarraysthatcanpotentiallyestimatebearingsto,

orlocalise,animals.


To identify which system is suitable for which kind ofmitigationmonitoring requirements, the operational

aspects of a platform need to be evaluated in combination with sensor criteria. Some of the points to be

consideredarealsoimportantforpopulationsurveysbut,ashasbeendiscussedinsection7.1,therearealso

someverydifferentrequirementsbetweenpopulationsurveysandmitigationmonitoring,soseparatesetsof

evaluationcriteriawereconstructed.

Themostbasicrequirementsformitigationmonitoringarethatdataretrievedbythesensormustbeavailable

innear real-timeand thatdetectionefficiencyover themitigation zone shouldbe sufficientlyhigh that the

chanceofmissingananimalislow.Eachdetectionmethodologyhasitsadvantagesanddisadvantages,e.g.PAM

canonlydetectvocalisinganimalswhileanyopticalmethodreliesonanimalsnearorbreakingthroughthesea

surface.Therefore,thesystemclassofthesensors(e.g.AAMorPAM)isthemostimportantcriteriawhenit

comestowhichmarineanimalspeciesactuallycanbedetectedandinwhichsituations.Thishasbeenintensively

discussedinVerfussetal.(2016).Forevaluatingthedetectionefficiency,thecollecteddatatypeandrelated

dataprocessingisimportanttoknow.Thesystemclassaswellascertaintechnicaldetailsofthesensorwillalso

determinewhichspeciesorspeciesgroupsareabletobeclassified.Todeterminewhetherananimalisabout

toenterorispresentintheexclusionzone,localisationoftheanimalinrelationtothesoundsourceneedsto

bepossible,whichcanbedonebyevaluatinginformationonbearinganddirectorhorizontalrangeofanimal

totheplatform(withpossibleambiguityregardingbearinganduncertaintyontheanimal’sswimmingdepth

whereonlydirect range isestimable)orat least thedeterminationof therangebetweenanimalandsound

source.Furthermore,itislikelythattherewillbeassociateduncertaintywiththelocalisationorrangeestimate.

47


DATE:July2016


Waysinwhichtominimisesuchmeasurementerrorneedtobeconsideredatthesurveyplanningstagetomake

surethatthemeasurementsaresuitablyreliableforuseinmitigationdecisions

Unless highly reliable automatic detectors are running on the autonomousplatform, itwill be necessary to

transferlargequantitiesofunprocesseddatainnearrealtime.Ifautomaticdetectorsareusedthenitwillstill

benecessary to transfer sufficientdata forhumanverification inorder toavoid falsealarms. Itmayalsobe

necessarytodemonstratethaton-boarddetectorsarerunningwithsufficientefficiencysuchthatnoanimals

aremissed.

Themaximummissiondurationaswellasthefactorslimitingitsdurationisonelimitingfactorthatdetermines

which monitoring requirements a system can fulfil. Also, theminimum, maximum and typical speed of a

platform is important forunderstanding thepotential tocoveramonitoringareaanddesign,aswellas the

environmentalfactorslimitingtheplatformperformance.

Thepayloadcapacity(seesection8.1.3)aswellasthetypeofinterfacedetermineswhatkindofsensorfitsonto

theplatform.Thatdeterminesmanyotherfactorsimportantformitigationmonitoring:Highnoiselevelscreated

bytheplatformorsensormayinterferewiththedetectionprobabilityorhaveaneffectonanimalbehaviour.

Thedatarelaysystemmayinfluenceifreal-timedetectionsareatallfeasiblebut,togetherwithitstransmission

range,alsodeterminesthemaximumrangeasystemcanoperateat.This,inturn,determinesthesizeofthe

monitoringareathatcanbecovered.Theverticalrangesaplatformcancoveralsodeterminesthedimensions

ofmonitoringareacoverage,whichforsomesystemshavetobecombinedwithcertainsensorpropertiesfor

meeting the requirementsneeded formitigationmonitoring.Forexample,ahigh flyingUASwillneed tobe

combinedwithahigh-resolutioncameratoretrievequalitativesufficientdataforanimaldetection,whilealow

flyingUAScanbecoupledwithasensoroflowerresolution.

8.1.3 Operationalaspects

Selectingthemostappropriateplatformandsensorsolelybasedonthecriteriaoutlinedinsections8.1.1and

8.1.2forconductingacertainkindofmarineanimalmonitoringisnotenoughtoproduceasufficientlyrunning

system.Therearefurtheroperationalcriteriathatarenecessarytoconsider.Forinstance,mostplatformswill

havesomemaximumpayloadspaceandpoweravailability,whichinformsastowhichsensorsmayfitintothose

payloadspacesandhowlongtheymightoperatefor.Therefore,definingthepayloadpower,capacityweight

andspaceforboth,platformandsensor,allowsanassessmentofwhichsensorsystemsaplatformcanfeasibly

carry.Bypayloadpowerwemeanthepowersupplynecessaryforagivensensorpackagetooperate–though

thisbalanceislikelytobeconfigurabletospecificmissionrequirements.Capacityweightandspaceidentifies

thenecessaryphysicaldimensionsandlimitationsofasensor/platformcombination.Thismayhavetwoaspects:

thespacewiththeplatformandthatpositionedoutside.Forsensors,thepayloadcapacitywasdividedintotwo

furtheraspects:theinternalpayloadcapacity,whichisthespacethatwouldberequirediftheelectronicsis

removed from the standard deployment package and incorporated into the vehicle payload bay, and the

48


DATE:July2016


externalpayloadcapacity,whichisthespacerequiredforthestandardmanufacturespackage.Operationallyit

isnecessarytounderstandthedeploymentandrecoveryproceduresfortheplatformandsensor,aswellasthe

manningrequirementsfordeploymentandrecoveryandtheleveloftrainingneededinordertobeableto

deploy,recoverandinterpretdatafromthesensors.Withdeploymentandrecoveryproceduresitisimportant

tonoteany requirements in termsofnumberof crewmen formanualhandlingand theneed forany lifting

equipmentatsea.Specialisttrainingmayberequiredfordeployment/retrievaland,morepertinently,operation

ofthecraft–potentiallyremotelyfromanonshorebaseaswellaspersonnelrequirementsforinterpretationof

data. It is also important tounderstand the failsafemodes theplatform incorporatesand the transponder,

detect andavoid capacities theplatformhas.Wedefine the failsafemodeas thedefault functionalityof a

system that allows it to operate shouldmajor problems emerge. This is of particular importance regarding

navigational and positioning ability within a field of industry operations. Failsafe mode options therefore

encompass:Bearing,todefinetheheadingofthecraft.Location,todefineitsgeographicposition.Endpoint,

notingitsfinalwaypointforretrieval.Keeptrack,describingthecoursetobeadheredtoorstop.Worstcase

scenarioswillhave tobeconsideredandtheabilityofaplatformto“limphome”andavoidbeingaserious

hazard requiresclose scrutiny.Considerationof transponder,detectandavoidcapacities takes intoaccount

abilitiesofsensor/platformtosafelynavigateandrespondtoitssurroundings.Furthercriteriarequestedwere

fuel type for the platforms and their level of autonomy. Fuel type varies greatly between platforms, and

individualcraftmayhaveanumberofoptions.Thisishighlyrelevantwhenconsideringmissionduration,safety

andmeansofre-fuelling/recharging.Understandingthelevelofautonomyiskeytothesuccessofanygiven

ASV/AUV/UASifitistoproveworthoverandaboveexisting“manned”methods,particularlyintermsofability

tomeetwaypointsandremainundercontinuouscontrol.

8.1.4 Datarelay

Datatransferrequirespowerandoftenalsoinfrastructuresuchasmobilephonemastsorsatellites.Offshore,

wheremobilephoneinfrastructureisgenerallyunavailable,theprinciplemethodsofdatarelayarecoveredby

thefollowingdatarelaysystemclasses:

• Wirelessmodems.Thesevaryfromsomethingsimilartohomeorofficewifi(highdatarate,butshort

range) to systems which have ranges in the tens of km, but lower bandwidth. They are generally

operatedaspeertopeernetworksandincurnodatatransmissioncosts.

• Satellitesystems.Thesecomeinmanyvarietiesfromsmall,lowpower,butalsolowbandwidthdevices

up tophysically large, highpower andhighbandwidth systems. These relay systemsare relianton

satelliteinfrastructureandthereforeincursignificantusagecosts.

• Analogsystems.Thesearerelativesimpleandasomewhatdatedtechnology,butcanbeusedtorelay

audioandvideodataovershortranges.

49


DATE:July2016


Thedatabandwidth isoneof themost fundamentalparametersgoverningthechoiceofdatarelaysystem.

Generally,bandwidthisquotedinbitspersecond,orbps.Mbpsmeansamillionbitspersecondandkbps1,000

bitspersecond.Notethatmostofustendtothinkaboutdataintermsof“bytes”,e.g.“Mymemorystickhas

an8GBytecapacity”.Thereare8bitstoabyteandthisneedstobetakenintoconsiderationwhenevaluating

thesedata.

For instance, a vehicle collecting long term monitoring data far from shore will have little choice but to

communicatethroughsatellitelinkanditmaybeacceptabletoreceiveeithernoreal-timedata,oronlyvery

basic summary informationon the systemstatus.Thesamevehicleandsensorcombination,whenused for

mitigation,mightoptforamuchhigherbandwidthradiocommunicationsystemtonearbyships,whichwould

allowmuchgreaterreal-timethroughputatnegligiblecost.

Theotherfundamentalparameterisrange.Transmittingdataovergreaterrangesrequiresmorepower.Allof

usareusedtoofficeandhomeWiFisystemswhichhaveexcellentdatathroughput,ofteninthehundredsof

Mbps.Goingintothenextroomcausedataratestodropanddisconnectionwilloccuratrangesinthelowtens

ofmetres.Datadelaywillgiveanunderstandingonhowlongoneneedstowaitfromsendingdatatoreceiving

it.Anothercriterionispowerconsumption:themoredatayouwanttosendandthegreaterthedistance,the

more power you’re going to need. For small, low power autonomous systems, power consumption can be

critical.

Clearly,thetransmitterunitneedstophysicallyfitwithintheautonomousplatform.Thisiswhytheoperational

sizeisimportant.Also,weightisanissuethatishoweverlessimportantforsurfaceandunderwatervehicles,

butessentialinformationifsystemsaretobeincorporatedintoautonomousaircraft.Platformrequirements

thatmightbenecessarytoinstalladatarelaysystemsalsoneedtobeconsideredaswellastheknowledgeon

anytrainingneedsnecessarytorunthesystems.

Fortheeconomicside,itisimportanttoinvestigatethepurchasecostoftransmittersandreceivers,aswellas

theiroperationalcost.Forsomesystemsusingfreespectrum,whereyoubuyyourownreceiver,operational

costsarezero(apartfrompower).Forsatellitebasedsystems,datacostscanbeconsiderable.Packagedsize

andweightofthesystemmayalsobeimportantforoverviewingupcomingcostsfortransportationifneeded.

Potential usage restrictions are also important to investigate as some radio based technologies may be

restrictedincertainareas/countries.

Environmental limitations to optimal system performance such as current or weather (wind, rain) were

requestedtounderstandinwhichsituationsperformancemaydrop.

8.1.5 Furtherinformationgathered

Generalinformationrequestedforthecomparisonmatrixincludethetypeofplatform,platformnameandthe

manufacturer and/or the designers. Next to the evaluation criteria as outlined above, further criteriawere

50


DATE:July2016


specifiedthatareimportanttonoteforanymonitoringandprojectplans, influencingthetimeplanning,the

budgetorallowingforadditionaldatatoberetrievedwhileperforminganimalmonitoring.

The technology readiness level as defined in Table 10 is important to assesswhether or not the system is

available for commercial use or its current stage of development. It was also prudent to determinewhich

platformsthesensorshavealreadybeenintegratedinto,orwherethereareplanstointegratethem,asthiswill

impactontheleveloftestinganddevelopmentrequiredforuseinautonomousmonitoringabilitiesinthenear

future.

Operationalandpackagesizeandweightoftheplatformsareimportanttoknowintermsofspaceandcostof

shippingaswellasspaceneededonasurveyvessel.Thisinformationisalsorequestedforthesensorsanddata

relaysystemstounderstandrelatedshippingcostsandimplementationpotentialintoaplatform(asmentioned

above).ThehazardclassofplatformandsensorasdefinedinTable11areespeciallyimportantforshippingand

transport issues. Here the hazard class levels related to battery or fuel type of an evaluated system was

evaluated.Onehastobearinmindthatoilsorothermaterialsusedinthemanufacturingofsystemsmayalso

havehazardlevelsassociatedwiththem,butthesecouldnotbeconsideredinthisreview.

Thepresenceorabsenceofadditionalsensorsmayhelptogatheradditionaldatabutmay,ontheotherhand,

alsobeafactorlimitingtheperformanceoftheplatformorsensorusedforanimalmonitoring.Oneotherfactor

interestingtoknowisthedeploymenttypeoftheCTDsensorifpresent.Thissensorcanbemountedeitheras

afixedsensororprofilingsensoronanAUV/ASV.AfixedsensorcanonlymeasureCTDdataatthelocationof

thevehicle.AprofilingCTDsensorcanmeasureCTDprofileofthewatercolumn.Thisisofparticularlyimportant

forASV,whereCTDprofilesofthewatercolumncanonlyberetrievedwithaprofilingCTD.

Forprojectplanningpurposes,itisofcoursealsoimportanttoinvestigatethecostsrelatedtothepurchaseof

theequipment,oralternativelytherentalprice.Operationalcostsandmaintenancecostsarealsoconsidered

forplatformandsensor.

8.2 Evaluationresults

8.2.1 UASplatforms

There are a number of different types of UAS,whichwe divided into powered aircrafts,motorized gliders,

lighter-than-air aircrafts and kites. These aredescribed in detail in the introductorypart of this review (see

section7.2).Forthisreview,weincludedpoweredaircraftsandmotorizedgliderswithfixedwingandrotary

wing-typesystemswhichenableoperationsfromlandandvessels,andallowforenoughpayloadtobecarried

for conducting animal surveys offshore for long term monitoring and real-time detection. Lighter-than-air

aircraftconsideredhereincludetetheredandremotelyoperatedballoonsandblimps,respectively.Tethered

balloonsarealsooccasionallycalledblimps,but,inthecontextofthisreview,thesewillbekeptseparate.Only

kitesthatcanberemotelyoperatedwereincludedinthisreview.Forthesensors,thosenotexceeding2kgand

51


DATE:July2016


able to provide an image resolution that allows for the detectionof animalswith at least 1m lengthwere

considered.

Basedontherecommendationsof(Koskietal.,2009a),wedefinedasetofcharacteristicsrelevanttowildlife

monitoringoffshoretounderstandwhichUAStoincludeintothisreview.Flightdurationandcommunication

rangewiththegroundstationshouldbesufficientlylongtoallowfordeploymentsfromlandandfromvessels.

Wedefinedthatdetectionofobjectsassmallasonemeterinlengthshouldbepossible,whichincludessharks,

turtles, fish shoals, and smaller marine mammals. This will allow UAS to be versatile in any environment,

regardless of the species present in any region. Additionally, to include both timely decision making and

populationsurveys,weconsideredaminimumhorizontaloperationalrangeof500mfromvesselsand200km

from land, considering the JNCC definition of high risk zone in marine mammal monitoring for industrial

operationsoffshore.ThelandrangeforUASoperationsdefinedhereallowsforfuturecomparisonstudiesto

determinewhetherthesesystemscanperformsimilarlytomannedaerialsurveys.Thisalsorequiresthatthe

flighttimeshouldbeatleast30minutesiftheUASisoperatedfromavessel,and4hoursifoperatedfromland.

Duetothe largeamountofUAS,wefocusedthecomparisonofUASsystemstothosethatcanbedeployed

rapidlyandwiththeneedofnomorethan3to4personstooperate.Theseareallintheclassofsmalltomedium

sizeUASandwithamaximumtake-offweight(MTOW)oflessthan80kg.Weconsidercharacteristicsthatare

relevantforanimaldetectionanddeploymentswithbothspaceandtimeconstraints.

The selection of platforms included different range, payload and deployment capabilities,whichmay serve

differentpurposesdependingonthestudiesofinterest.Weincludedsystemsthatcanbeappliedforpopulation

andmitigationmonitoringaswellasmonitoringofindividualanimalsfortheapplicationintheO&Gindustry.

8.2.1.1 Poweredaircrafts

PoweredplatformshavebeenmostwidelytestedintheUASfield.Thecapabilitiesofthesesystemsallowthem

tobeconsideredasalternativemethodsforworkdevelopedbothatseaandonland.Dependingonthestudy

design, there is currently a wide variety of platforms available at the moment that can overcome many

limitations that human observers and other aerial platforms (e.g. digital surveys using manned aircraft)

experience. The versatility of these systemshighlights thepotential applicationsof this equipmentboth for

researchandmanagement.

Inmostcommerciallyavailablesystems,themanufacturerisabletoprovideafullyintegratedpackagethatcan

satisfytheneedsofoperatorsandclients.However,listedsensorsystemsinthecomparisonmatrix(Appendix

11.3.8)havesimilarcharacteristics,especiallyforthermalIRandRGB(RedGreenBlue)cameras,whichmaybe

applicabletoplatformsthathavegimbalcapacity.

NeitherofthelargestUASmanufacturerslistedinthisreport,suchasInsituandArcturusUAV,providedaprice

ontheirsystemuponrequest.Specificrequestsforpriceestimatesorpricerangeswereunsuccessful.Inother

occasions,quotesweregivenonaspecificsystemwithdetailsregardingthequote,suchasprice,beingundera

52


DATE:July2016


strictNon-DisclosureAgreement. Themilitarymarket clearly is thebiggestmarket for these systems today,

thereforekeepingthepriceofsuchsystemsasecretmaybebeneficialwithregardtopricedifferentiationamong

differentcustomers.However,accordingto(Thammetal.,2013),thesesystemstendtovaryinpricebetween

US$27,360toUS$76,577.Systemsbasedonmilitaryapplications,suchastheScanEagleforinstance,mayreach

highercostsduetothedemandformorecomplexlaunchingandlandingequipment,andcontractswithcertified

operators.

UASplatforms are generally very versatile.Dependingon the studydesign andmain goals of a survey, it is

possible touseawidearrayof systemsthatcanprovidedifferentdata types.Twoof themainaspects that

shouldbeconsideredareflightdurationandrangefromthegroundstationandthemainpilot.Thispointalone

willhelpdefine theUAS that suitsa study in thebestwaypossibleand fromthere, restrict theselectionof

sensorsthatareapplicable.UsingIridiumsatellite,mostsystemsconsideredcanbepilotedglobally.However,

forlivedatastreaming,thepayloadcanbelimited(unlesstheoperatorhasaccesstoasatellitedatalink,orcell

phonecoverage).Thechoiceofdatalink,andhencerangevariesdependingonlocation,asregulationsregarding

frequencybandsandtransmitterpowervariesfromcountrytocountry.

MostUASareabletoconducttransect-basedsurveys.Bothfixed-wingandVTOLplatformsareabletoconduct

waypointflights(transects)andfocalfollows.However,fixed-wingplatformsmayhavedifficultiesinhovering

andrelyongimbalsensorstolockthecameraonthepointorobjectofinterestwhileloiteringaroundthetarget.

The choicebetweendata in the formofphotoor videowill alsobedependenton the study’sgoal and the

resolutionthatisrequired.However,aswithanyothervisualmethodofobservation,UASareonlycapableof

detectinganimalsthatarevisibleatthesurfaceorinthetoplayersofwater,whichmayprovetobeefficientfor

studiesconcerninganimalsthatspendlongperiodsoftimeatthesurface.Foranimalswithlongdiveduration

(e.g.beakedwhales),theflighttimerequiredismuchgreaterandstudiesforthosespecieswouldprobablyrely

onacombinationofsurveymethodstoaiddetection.AsUASevolve,sodotheircarryingcapacity(payload),

whichcould include thedeploymentofacoustic technologyduring flightwhileat thesame timeconducting

visualsurveys.Thecombinationandcomparisonofdifferentsystemsisatopicwhichhasbeenunderdiscussion,

asallcurrentmethodshaveadvantagesandlimitations,andmayprovidemoreaccurateestimatesofanimal

positioning,size,andmoredetaileddescriptionsofanimalbehaviourinanimalstudiesatsea.

TheonlytwofullelectricpoweredaircraftproposedinthisreviewaretheTrimbleUX5andtheBramorC4EYE.

These systemsbenefit fromthe lightweight,easeofuseand the lowoperational cost, requiringonly1 to2

personsforoperating.Theybothcanlandinconfinedareasandarewellsuitedforshortmissionsclosetothe

shoreline.TheBramorC4EYEhasthelongestflightduration(uptothreehours)andhasalsothecapacityfor

streaminglivedatabacktotheoperator,whilethedatahastobedownloadedfromtheTrimbleUX5afterthe

missionhasended.Fourfuel-poweredaircraftsareproposed,wheretheArcturusJump-20istheonlyaircraft

withVTOLcapabilities.Thisisaclearbenefitwhenoperatingfromshipsorinareaswithoutasuitablelanding

strip,andrequiresnoextrainfrastructureotherthanaflatsurface,suchasahelipad.TheInsituScaneagleand

53


DATE:July2016


ThalesFulmarbothhavecapabilitiesforwireornetlandingonshiporland.ThePenguinBistheonlyaircraft

thatrequireanairfieldtoland.Withdurationof24hoursithasalongrange,andshouldbeconsideredincases

whereonecanoperatefromanairfield.TheThalesFulmaristheonlyproposedaircraftwhichcanlandonsea

andiswellsuitedfortheharshenvironmentofmarineoperations.

Withmore than800,000proven flighthours5 the Insitu ScanEagle is in a classof its ownwhen it comes to

mediumsizedUAS. Ithas thecapabilities tooperate fromshipsandruggedterrainandhaspreviouslybeen

evaluatedforsurveyingofmarinemammalsandmarinefauna.Withamissiondurationof24+hoursandits

capacityforstreaminglivedatabacktothegroundstationitcanbeusedforcontinuouslymonitoringoflarge

areas.However,apossiblelimitationforallcurrentlyavailablemediumsizeUASisthelackofanintegratedde-

icingsystem.

Inpoweredaircraft,thedatacanbestoredintheformofSDcardsandthendownloadedtoalocalcomputer

foranalysis.Imagesandvideomaybeeditedaposterioriandprocessedusinganalysissoftware(Irelandetal.,

2015) and human visual inspection. Live streamingmay also be possible,with storage and further analyses

occurringatthereceivingend.TheonlyproposedsystemwithoutthiscapabilityistheTrimbleUX5.

8.2.1.2 Gliders

Gliderplatformstendtobe lessversatile thanpoweredaircraft,giventhattheyaredesignedtoharvest the

energypresentinrisingair.Hence,theydonotprioritizemaintainingaltitudeorflyinginstraighttrajectories.

Further,thecapabilitiesarethesameasfixedwingpoweredaircraft.Thesesystemswerenotfoundsuitablefor

marineanimalstudiesduetothesinkzoneabovetheocean.Hence,noglidersystemswereconsideredinthis

evaluation.

8.2.1.3 Kites

Kitesaregenerallylimitedtowindspeedsunder6ms-1,whichmaylimitthelocationsinwhichtheycanoperate

(Thamm et al., 2013). Adding to the dependency in environmental conditions, this equipment is not often

stabilized, resulting indifficulties inhoveringoveranobjector locationof interest.Thenumberofavailable

platformsislimited.Often,parafoilkitesmaybeusedandthepartscollectedtocreateasystemusedforsurveys.

Thereareonlineplatformsavailablewheretheconstructionof thesesystems isadvisedandguided,making

theseplatformshighlypersonalised.Thesesystemsarenotwellsuitedformarineanimalstudiesduetotheir

wind limitations.Thereare fewcommerciallyavailable systemsor systems thatarebecomingavailableasa

completepackage,andwehavethereforeselectedthesinglekiteplatformthatisabletoconductsurveysof

marineanimalsinflexibleweatherconditions.

5 www.insitu.com

54


DATE:July2016


BasedontheestimatesbyThammetal., (2013), it isestimatedthatthepricerangesforkitesystemsrange

between~US$20,000and~US$44,000.

Kitesaresimilartofixedwingpoweredaircraft,exceptthatthewingsarenotrigid.Theadvantagesare

• Easiertocarryinabackpackthancomparedtoafixedwingwithsamepayloadcapacity.

• Neednorunwayfortakeoffsandlandingsandcanbeoperatedfromaship.

Thedisadvantageistheextremely lowweathermaximaandlimitedcontrolduringflight,whichmakesthem

unusableforoffshoreapplicationssuchasmonitoringofmarinemammals.

Kitesaremoreweather-dependentthantheothersystemsabove.Theyareparticularlysensitivetowindand

rain,asthesecanaffectthestabilityoftheequipmentleadtodifficultiesinmaintainingaspecificsurveytrack.

These UAS may conduct transect-based surveys for animal population studies, but struggle in maintaining

hoveringpositiontoperformmorefocalassessments.Windspeedmayhaveastrongeffectonkitestabilityand

alsoontheamountofoverlapthatmayresultfromasurvey,andshouldthenbetakenintoconsiderationduring

flightpreparations.AswithothertypesofUAS,itispossibletochoosebetweendataintheformofphotoor

video, and as any other visual detection methods kite systems maintain the same limitations concerning

detectabilityofanimals.Thepayloadcapacityofthesesystemswilldependonthesizeofthesystem,thoughit

hasyettobetestedatsea.

Kitesystemshaveaflighttimethatcangouptoafewhours,thoughitseemstorequireslightlylesstraining

timeforpilotsandoperators(Thammetal.,2013).Similartopoweredaircraft,thesizeofthekitewillprovide

an indicationofthepayloadthat itcancarry,thoughsmallerkitesmayprovetobemoreunstable instrong

wingsthanlargerones.Theseplatformsrequirerunwaysfortake-offandlanding,thusmakingthemdifficultto

operatefromvesselsinareasdistantfromthecoast.Seesection7.2.3fordetails.

Forthesesystems,thedatacanbestoredinthesameformatasforthepreviousaircraft.Imagesandvideomay

bealsoeditedaposterioriandprocessedusingbothanalysissoftware(Irelandetal.,2015)andhumanvisual

inspection.

8.2.1.4 Lighter-than-airaircraftsystems

Similartokites,thesesystemsmaybedependentonweatherconditionstoprovideadequateimagesofthearea

surveyed.However,newsystemsarebeingdevelopedthatmayovercomethislimitation.Wethereforeinclude

theseplatformsinthelistofmethodsformarineanimalmonitoring.

TheOceanEye is suppliedwith a triple sensor unit (EO/IR/AIS) and is capable of real-time, night video, and

imagery.TheelevatedAISreceiverallowsforincreasedshipdetectionrange,whichisofvaluewhenconducting

studiesofanimalswithastrongperceptionofvesselpresence.Thecompactcameraisattachedtotheballoon,

improvingthestabilityofthesensorunitandqualityofthefootage.TheHelikiteplatformisabletocarryGyro-

55


DATE:July2016


stabilisedandstandardPTZ (pan-tilt-zoom)andmodularvideocameras,whichcanberemotelyoperated to

provideawiderview,orstoreinformationon-board.Bothcanbeattachedinanypositionofthetetheringline,

allowingforanexchangeofbatterieswithoutlandingtheballoon.

NopriceinformationwasrequestedfromthemanufacturerfortheOceanEyesystem.Basedonthepriceranges

oftheHelikiteweestimatethatthepricerangesforballoonplatformsrangebetweenUS$147andUS$7,362.

TheseUASmaybelessweather-dependentthankites,astheyaregenerallytetheredtoavessel.Theyarealso

sensitivetowindandrain,butareabletomaintainacertaindistancefromthesurveytrackorthestudyanimals

duetothepresenceofacablethatstabilizes(uptoalimit)winddisturbance.TheseUASmayconducttransect-

basedsurveysforanimalpopulationstudies,andcanbecontrolledtomaintainahoveringpositionandperform

morefocalassessments.However,thisequipmentisdependentonthepresenceofthesupportvessel,which

mayaffectanimalpresenceandthuscreatebiasedstudies.

Tetheredballoonscan,atstandardtemperatureandpressure,carryaboutafewkilogramsofpayloaddepending

mainlyontheirhelium(orhydrogen)capacity.However,thoughnotgenerallyacknowledged,theweightofthe

required length of linemust be taken into account tomake accurate assumptions about the actual sensor

capacityoftheballoonorblimp.Additionally,weatherconditions,particularlyheatandhumiditymayaffectthe

lift of the balloon,which in conditions of rain, snow, andmist can get extraweight from the accumulated

precipitationon topof theplatform.This typeofplatform isgenerally sensitive towindconditions.Helikite

managedtoovercomethislimitationbyincorporatingatetheredheliumballoonwithakitewing.Furthermore,

giventhattheyaretethered,thelengthofthelinecanbeadjustedtocoverdifferentheightsanddistancesfrom

theattachedvessel.

Inlighter-than-airaircraft,thedatacanbestoredon-boardtheaircraftanddownloadedtoalocalcomputerfor

analysisafterthesurveyhasbeencompleted,ordirectlytransmittedtoagroundstationpresenton-boardthe

supportingvessel.Throughreal-timetransmissionofvideoimages,itisalsopossibletotriggeracameratoselect

photosofthetargetobjects,ratherthanstoringthewholevideo(Hodgson,2007).

8.2.2 UAS-sensors

Fromthediscussionabove,welimitthesensorsurveytostabilizedcamerasystems(i.e.gimbalsystems)capable

ofacquiringHDvideo(orbetter).Mostgimbalsystemsonthemarketcomepre-assembledwithsensors,and

themanufacturers gives the customer a choice froma selection of different sensors to suit the customer’s

application and needs. As the systems listed below are expensive, high performance gimbals, most

manufacturersdeliverthesamecamerasensors,orcameraswithsimilarcharacteristics.Hencethechoiceof

gimbalsystemoftencomesdowntowhattheUASsystemprovidercandeliverasstandardoptions.

Thebiggestgimbals(likeCloudcapTASE400)areonlyneededforexotic(combinationsof)cameras,suchasthe

VIS+LWIR+SWIR,orforheavyopticalzoom(whichweprobablydonotneedformarineanimals,aswemost

oftenwishtoflybelow120metersduetoaviationrequirementsforVLOSoperations).

56


DATE:July2016


• CloudcapTechnology:Tase400&Tase310

• UAVVision:CM202&CM100.TheCM100Gimbalisthesmallestandlightestinthestudy,henceitis

possibletofitmostUASsystems

• DST:OTUSU135HIGHDEF&OTUS-L205HIGHDEF.

Inaddition,therearespecializedgimbalsfromInsituandC-ASTRALwhicharecompatiblewiththeScanEagle

andBramorsystems,respectively.

Mostplatformsconsideredinthisevaluationareavailableasaset.Itispossibletoexchangethesensorsfora

singleplatformbutitwilldependontheavailabilityofcompatiblesensorsdevelopedbythesamemanufacturer.

ItmayalsobepossibletodeploysensorssuchasDSLRcamerasasgiveninthestudyexamplesofTable7,butit

hastobetakenintoconsiderationthepayloadcapacityoftheplatformanddatatransmissioncompatibility.It

isthereforerecommendedtofollowmanufacturers’adviceonthepossibilityofusingalternativesensors.

Giventhatforaturn-key(fullyintegrated/assembledandready-to-go)UASsystemthesensorsandstabilizing

gimbalarepurchasedasapackage,andthe informationpublicallyavailable is limited, itwasnotpossibleto

obtainpricerangesforsinglesensors.

Imaging sensors suchashigh resolution still or video camera, operating in the visible spectrumor thenon-

thermalregionoftheinfraredspectrumarethemostrelevantsensorsformarineanimalstudies.Forvideo,itis

importantthatthecameraisgyrostabilised.Thiswillimprovetheefficiencyofoperatoranalysingthedataand

willalsoreducetherequiredbitratewhendoingvideoencoding/compression.Itisalsoimportantthattheimage

framesaregeocoded,sothatthelocationandsizeofaspottedanimalcaneasilybedeterminedfromtheimage.

Afuturesystemcombininggyrostabilisedgeocodedvideowithcorrespondinghighresolution,geocodedand

overlappingstillimagesseemstobeaneffectivesystemfornear-real-timedetectionandidentificationofmarine

animalsfromUASdata.Insuchasystem,onecouldimaginetheoperatoridentifyingpotentialsightingsinthe

video stream by point and click, and using the corresponding high resolution still image from this area for

verifyingandidentification.

UASsensorsarepositionedbelowthesurveyplatformeitherprovidingastraightorangledview.Thechoiceof

video(real-timeoron-boardstored)orstillimage,andgroundresolutionwillrelyontheobjectivesofthestudy,

butthechoiceofthetypeofsensorhastobetakenwithconsiderationforthepayloadcapacityoftheUASto

beused.SomemanufacturerssuchasARCTURUSUAVandTRIMBLEprovideonlycompletepackages,inwhich

case one can only operate using the sensors provided in the package or produced specifically by that

manufacturer.

AnintegratedsystemforanalysingthedatafromtheUASisrequired.Somedataprocessingcapabilities,such

asobjecttrackingormotiondetection,canbeincludedinthesensororgimbal.Thishasmostlybeenusedin

systemsdesignedforsurveillancepurposes,andmaybeusefulformarineanimalmonitoring,especiallyforfocal

followsandmitigationpurposes.Thecommunicationsystem,usedfortransferringdatafromthesensorandto

57


DATE:July2016


thegroundisoftenanintegratedpartoftheUASandalsousedforthemainorsecondarytelemetrylinkfor

monitoring and control of the aircraft. Depending on the aim of the study, it is possible to createmosaics

(overlapping the photos and creating a chart), analyse each photo or video individually, or for real-time

detection,usingasystemsuchastheonedevelopedbyIrelandetal.,(2015).Giventhatthedatamaybestored

atalltimes,therewillbenodatalossesunlessthesensorisnotproperlyconnectedtotheaircraftorthereis

equipmentfailure.Analysingeachphotoorvideoindividuallycanbequitetimeconsuming,andwetherefore

urgethedevelopmentandpublicavailabilityofdetectionalgorithmsthatcanbeusedtotriagethephotos.

8.2.3 AUV/ASVplatforms

8.2.3.1 Propellerdrivenunderwatercraftandpoweredsurfacecraft

As ismade apparent in thematrix given in section 11.4, the range of potential combinations is extensive.

Dominantfactorsinshapingoptionsarethepayloadsizeandpowerbudget.Missiondurationandwhetherthe

data istransferredorrecordedareearlyquestions indefiningtheexactspecificationsforanygivenmission.

Whilsta“lowcost,off theshelf”approach is slowlyemerging (more rapidly inAUVthanASV), inalmostall

instancesspecificrequirementscanbetailoredtoaparticularsurvey.Onthepracticalside,self-noiseisakey

issue to be addresses by all crafts if applied to PAM or AAM. The craft itself may be very quiet but the

deployment/attachmentmethodofacousticsensorisallimportanthere.Self-propelledAUVhaveaparticular

challengeinthis.Alsotonote,thatcameras,bothHDandIRcanbeattachedtoASVwhich,thoughitoffersa

very low observation platform, may have an application in the detection of surface animals, e.g. marine

mammals,turtles,seabirds.

The purchase price range of the above AUV and ASV for existing models with integrated sensor packages

includedspansfromtensofthousandsofpoundsforsmallerAUVtoseveralhundredthousands,evenmillions,

ofpoundsforhigherspecificationinstruments.Anothermodelisrental.Thisisusuallybythedayandcostshere

areestimatedtobelowhundredsofpoundsperdayforlowerspecificationAUVuptoseveralthousandperday

for more. Pricing, especially for rental, may be subject to deployment length and exact requirements.

Manufacturers typically have been open to integrating new sensors and tailoring specifications for a given

project,butthisincursresearchanddevelopment(R&D)costs–whichcanbesignificant.Itisworthnotingthat

asthistechnologyemergesoutofR&Dlaboratoriestoseekanentranceintothecommercialmarket,anexact

“pricepoint”oftheproductisoftenstilltoform.Additionally,associatedcostssuchastransportationandservice

arrangementsshouldbekeptinmind.

AllaboveAUV/ASVaredesignedtofollowatracklinebywaypointsetting.Variationemergeswithoperator

capabilitytoaltercourse.AllASVclaimthisability,butquestionsariseonresponsiveness.Iridium,freewave,wifi

andacousticmodemcommandandcontrolenablebuoyancygliderstobedirectedinreal-time.Asaresultthey

are targeted platforms, and can be used to investigate ambiguous observations, or to change survey areas

rapidlyifrequired.Station-keepingabilitiesvarywidely.ItisclaimedbysomeAUVbutthereisuncertaintyhere

58


DATE:July2016


as to exact capabilities. Powered ASV have the potential to stay on point more dynamically, but this has

implications on power/fuel. Themethod of staying in position by close circling hasmore flexibility but, by

definition,provideslesspositioningprecision.MoreclearlyknownonmobilityisthatallAUV/ASVmovevery

slowly.Except,thatis,forthepoweredboat-styleRIBs,someofwhicharecapableofveryhighspeedsindeed.

AllAUV/ASVaretheoreticallycapableofconductingPAMandAAMsurveys,butprojectspecificsmustbeborne

inmind.Datatransfercapabilitieswilldeterminewhetherthisisforpurposesofloggingdataonly,near-real-

timeorgenuinereal-time(i.e.formitigationmonitoring).Additionally,self-noisemaybeanissueinsome(such

aspropellerdrivenAUV).

WhenconsideringoperationalpracticalitiesofASV/AUV,aguidingthemeissafety.Anyriskof,forinstance,AUV

/ASVbreakdownaheadofaseismicvesselandassociatedequipmentwouldhavemassiveimplicationsforthe

wholeoperation.Beforeusageinindustrialfieldofoperations,extensivetrialsandproventrackrecordwillbe

requiredforallcraftstoprovideassuranceontechnicalreliability.Highconfidenceinabilitytocompletemission

durationandgenuineindependencearealsohighpriorityattributes.In-fieldsupport,suchasbyachasevessel,

may be available but not guaranteed and exact requirements will vary by application. Any ASV/AUVmust

demonstrate full capability in these areas. Logistical factors must be taken into account, such as ease of

deployment/retrieval.Safetyconcernsareparamount,butalsoimplicationsoncost(suchasfreighttooverseas

locations).

Regardingremoteoperation,technicalandsafetyassuranceswillberequiredonthedependabilityofthelink

andresponsiveabilityofremoteoperation.Thecontrol,handlingandmaintenanceofASVandAUVmayrequire

specialistknowledge,providedeitherbyadditionalAUV/ASVexpertsheadingoffshoreorbytrainingofexisting

crew.Aclearcodeofpracticeislikelytoberequiredanddisseminatedinadvancetopotentialmarine/airspace

users,includingthirdparties,inthearea.

8.2.3.2 Autonomousunderwaterbuoyancygliders

Thereare,toourknowledge,currentlyfourcommerciallyavailableelectricbuoyancygliders:theSlocumelectric

(Webb et al., 2001)manufactured by TeledyneWebb, the Seaglider (Eriksen et al., 2001)manufactured by

Kongsberg,theCoastalglider(ImlachandMahr,2012)manufacturedbyExocetus,andtheSeaExplorer(ACSA,

2013)manufacturedbyACSA.Inaddition,thereareseveralunderwaterglidersthatareinuseordevelopment,

includingSpray(Shermanetal.,2001),Deepglider(OsseandEriksen,2007),Tsukuyomi(Asakawaetal.,2011)

andtheLiberdadeXray/Zray(D'Spainetal.,2011;D'Spain,2009).TheCoastalgliderisdesignedforuseinthe

littoral zone (it is self-ballasting fromessentially fresh to fulloceanwater),witha fastermaximumspeed (2

knots)(ImlachandMahr,2012).TheDeepgliderisdesignedtoglidetodepthsof6,000m(OsseandEriksen,

2007),theSeaExplorerisdesignedtobefaster(1knot)andpoweredbyrechargeablebatteries,theTsukuyomi

isbeingdesignedforlongdurationasavirtualmooring(Asakawaetal.,2011)andtheLiberdadeZray/Xrayis

designedforlongdistance,longdurationcarryinglargeandhigh-data-ratepayloads(D'Spain,2009).Currently

59


DATE:July2016


thereisonemanufacturerofathermalglider(TeledyneWebb),andfourhybridvehiclescurrentlyadvertisedor

in-use,theSlocumhybrid(Jonesetal.,2011a)manufacturedbyTeledyneWebb(note,allnewSlocumG2gliders

have the hybrid capability), the eFòlaga (Alvarez et al., 2009)manufactured by Graal-tech, the SeaExplorer

(Rochet,2015)manufacturedbyACSA,andtheexperimentalSeaBird(ArakiandIshii,2007).Wehavelimited

thedescriptionsbelowtothoseglidersthatarecurrently(orhavebeen)commerciallyavailable.

Thefirstocean-provenelectricalbuoyancygliderswereSlocum,SeagliderandSpraywithsimilarfeatures(Table

1).Themorerecentdesignstypicallyfocusonparticularimprovements/applications.

Table1.Commondesignfeaturesofelectricbuoyancygliders(fromDavisetal.,2002andWoodandMierzwa,2013).

Feature Design

Endurance

Endurance ofmonths to year, at slow horizontal speed (~0.3m/s) to

minimisedragandverticalvelocitiesof~0.1m/s

Ballast system that uses a hydraulic pump to move oil between an

externalbladderandaninternalreservoir.

PortabilityandversatilityA similar size, shapeandweight (~2m length,~60kg inweight), that

permitdeploymentandrecoveryby1to3peopleandasmallrib

EconomyinuseConstruction costs in theorderof $50,000 to$100,000 and refuelling

costs$3,000to$15,000perdeployment.

Real-timecontrolanddatarelayGPSnavigation,on-boardPC-levelinternaldataprocessing,andanability

toreceivecommandsandtransmitdatatimely.

Inmost commerciallyavailable systems, themanufacturer is able toprovidea fully integratedpackage that

integratesapassiveoractiveacousticsystemintotheunderwaterglider.Selectionofaspecificsensormaybe

limitedbythevehiclepayloadsize (which issmallbutcomparablebetweenthegliders), thecommunication

methodemployedbytheglider,andsuitabilitytoaplatformundertakingasaw-toothprofile.

Thepricerangeforallunderwaterbuoyancyglidersspans~$50,000to$200,000forexistingglidersandpre-

integratedsensorpackages.Themanufacturershavetypicallybeenopentointegratingnewsensors,although

thesehaveincurredresearchanddevelopmentcosts.Newlithiumbatteries,sensorcalibrationandre-ballasting

for amission are in the order of £5,000 to £15,000. Piloting telemetry costs are ~£2 / day, although data

transmissioncostscanbesignificantontopofthis.

Buoyancy gliders typically undulate in the upper 1,000m of thewater column, therebymaking subsurface

measurements.Traditionalvisualsurveymethodsinvolvingaircraftorshipsareexpensiveandcanbeineffective

atdetectingsmallaggregationsofanimalsthatspendsignificantperiodsoftimesubmergedandoutofviewof

theseasurface.Visualmethodsarealsonaturallylimitedbyfactorsthataffectvisibility,suchasroughseas,fog,

rain,snow,anddarkness(Baumgartneretal.,2013).Thebuoyancyglidercanthereforeundertakemarineanimal

surveysinconditionsnotsuitableforvisualmethods.

60


DATE:July2016


Differentmakesofbuoyancyglidershavedifferentmissiondurationsbut typically, oncedeployed, theyare

operational 24 hours a day. Assuming themethod of detection does not require light, thenmarine animal

observationsareundertakencontinuallythroughoutthemissionduration(6hoursto>1year).

Iridium,freewave,wifiandacousticmodemcommandandcontrolenablebuoyancygliderstobedirectedin

real-time.Asaresult,theyaretargetedplatforms,andcanbeusedtoinvestigateambiguousobservations,or

tochangesurveyareasrapidlyifrequired.

Ships and aircraft are operationally expensive to run.With a purchase cost of ~£100,000 per platform and

running costs of ~£10,000 to £20,000 per mission (battery, piloting communication and CTD calibration),

buoyancyglidersareascalableeconomyforasurvey.Newmethodsforcontrollingmultiplevehiclesenablea

multi-vehicleapproachtosurveysthatmaycompensatefortheslow-movingnatureoftheplatform(Greene,

2014).

Buoyancyglidersareacousticallyquietplatforms,sincetheydonothavethrustersanduseinternalactuators,

makingthemidealplatformsforacousticmonitoring(bothAAMandPAM).However,acousticplatformnoise

(self-noise)generationdoesoccurs(predictably)whenthebuoyancychangeandtrimadjustmentsmechanisms

areactivated,causingpotentialperiodsofmaskingduringPAM(Ferguson,2010).

Buoyancyglidersaredesignedforslowspeedstoreducedragandextendendurance.Typicalspeedsare~0.3

m/s,with the addition of thrusters this can increase speeds to ~1m/s. The ability of a glider to perform a

transect,ortopassthroughawaypointisdependentontheoceancurrentsandtheirvariabilitywithinthesurvey

region.Theoperatorcanspecifyhowclosetoawaypointtheplatformshouldachievebeforeitisacceptedto

havereacheditsgoal,howevertheroutetothatwaypointcanbeconvolutedandstraight-linetransectsmay

thereforenotbeachievableindynamicoceanenvironments.

There is limited information on the reliability of gliders, due to their relative newness. Brito et al., (2014)

investigated the reliability of gliders using 205missionsmade by 56 gliders undertaken by the EUGROOM

(GlidersforResearch,OceanObservationandManagement6)programme.Theprobabilityofadeep(1,000m)

underwaterglider,independentofmanufacturer,survivinga90-daymissionwithoutaprematuremissionend

is approximately 0.5. The probability of a shallow underwater glider surviving a 30-day mission without a

prematuremissionendis0.59.

8.2.3.3 Self-poweredsurfacevehiclesanddriftingsensorpackages

In recent years, several self-powered surface vehicles havebecomeavailable. These vehiclesmostly extract

energyfromwavemotionandconvertitintoforwardmotion.Somevehiclessupplementthiswithsolarorwind

powered electricity generation which is then used to drive a propeller. By extracting energy from their

6 http://www.groom-fp7.eu/doku.php

61


DATE:July2016


environment,thevehiclescanoftenstayatseaalmostindefinitely.Electricalpowerforpayloadsissuppliedby

batterieschargedbysolarpanels.Control issimilar forallvehicles.GPSandon-boardsoftwarenavigatethe

vehicle between waypoints. Vehicle position and be monitored via satellite and new instructions can be

uploadedatanytime.

Self-poweredanddriftingsurfacevehiclesarerelativelyquiet,havingnopropellernoise.Thismakesthemhighly

suitable forPAMaswell asAAM.Someresearchershavealsoattachedcameras to thesevehiclesandhave

detected birds andmarinemammals. Purchase prices are in the $100,000’s for theWavegliders while the

smallestoftheMOSTAutonautvehiclescostslessthan$100,000.

Alloftheself-poweredsurfacevehiclescanfollowatrackline.However,incalmconditions,whenwave-power

isunavailable,theywillnotbeabletodosoaccuratelyandinflatcalmconditions,willdriftwiththecurrent.

TheexceptionistheSV3,whichhasapropellerthatcanbeswitchedonatthesetimes,solongassufficientsolar

powerisavailable.Duetotheirsub-seastructurewhichdescends7mbelowthesurface,wavegliderscanonly

beoperated inat least12–15mwaterdepth (shallower-water versions canbeavailableon request). This

generallyrequiresdeploymentfromavesselequippedwithasmallcranetolowerthefloatandsub-unitover

theside.Wavegliderscanbedeployedfromaslipifthesub-unitistieduptothefloatandthevehicleistowed

outtodeepwaterpriortothesubunitbeingreleased.

TheAutonautvehicleshaveasmalldraftandcanthereforealsobeeasilydeployedfromaslipwayorbeachas

wellasfromasmallvessel.Oncedeployed,thevehiclescanstayatseaformanymonthsandbothWavegliders

andAutonautsareknowntohavesurvivedseverestorms.Communicationwithshoreviasatelliteisprovided

on all Waveglider and Autonaut vehicles except for the smaller Autonaut 2 which only has UHF andWifi

communication. Where satellite links provide insufficient bandwidth for data transfer, it is possible to fit

additionalradiolinkstothesevehiclesforhighvolume/shortrangecommunication.

8.2.3.4 DataprocessingandtransferinAUVandASV

Data processing and transfer issues are generic to AUV and ASV platforms. Basic communication (piloting

telemetry)withvehiclesisundertakenwitheasethroughthecurrentsatelliteorRFmodemoptionssincedata

volumesarerelativelysmall.However,datarelaygenerallyrequiresconsiderablymorebandwidthandmaybe

impossibletoachieveunlessahighlevelofautomatedprocessingcanbeimplementedinordertoreducedata

volume.Challengesoccurdependingon the sensors fitted to thebuoyancygliders, andwhat information is

requiredtobetransmitted.Therefore,theseissuesarerelatedtothesensorfittedtotheplatform

8.2.4 AUV/ASV-sensors

Thereviewofsensorshasbeenlimitedtothosewhichhavebeendeployedonautonomousplatformsorwhich

performahighlevelofprocessing,makingthemsuitableforreal-timedetection.Thisreviewthereforeexcludes

mostofthePAMsystemsdesignedforfixedautonomousmonitoringreviewedinSousa-Limaetal.,(2013).

62


DATE:July2016


8.2.4.1 PAM

Todate,mostinterestfromresearchershasbeeninthelowercostsubmarineglidersandself-poweredsurface

vehiclesinthehopethattheywillprovidequiet,persistentplatforms,capableofcollectingPAMdataonlarge

temporal and spatial scales. The only PAM systems which are suitable for use on small lower powered

autonomousvehiclessuchasunderwaterglidersaretheSoundTrap,theDMONandtheWISPRboard.Indeed,

theWISPRboardisavailableasastandardpackagewiththeKongsbergseaglider.However,itsrelativelyhigh-

powerconsumptionsignificantlyreducesthelifetimeofgliderdeploymentsfrommonthstoweeks.

Thesurfacevehiclesthathavebeenreviewedcouldpotentiallyworkwithanyofthesensorslistedsincethey

havefewerspacerestrictionsandgenerallyhavemorepoweravailable.However,theremaystillbetrade-offs

betweenpowerconsumption,poweravailabilityandmissiondurationformanypossiblecombinations.

Thelargerpoweredvehiclescouldalsoaccommodatewidebandsatellitecommunicationssystemsandcouldin

principle,providemorereal-timemonitoringwhenoperatingremotely.

WeareunawareofanyPAMsystemsbeingusedontherangeofpoweredunderwatervehicles.Thereisnothing

fundamental preventing this from happening beyond the obvious possibility of noise interference from

propulsionsystems.Mostlikelyresearchershaveshownlittleinterestinthisduetotheshortmissionduration

ofthesevehicleswhichwouldmakethemimpracticalforthecollectionofmarineanimalsurveydata(which

generallyrequiresdatacollectionoverlongtimeperiodsinordertoobtainsufficientsamplesize).

Wheregiven,pricerangesforPAMsystemswereallinthe$2,000to$10,000pricerange.However,itshouldbe

notedthattheremaybeengineeringcostsassociatedwithmountinganyofthesystemsintospecificvehicles.

Theseadditionalcostsmaybeneededtocoverinstallationofthemonitoringdeviceintoasecuredryspace,

mountingofhydrophones,supplyofpowerandinterfacingofanycommunicationssystems.Thesecostsare,

however,likelytoremainlowcomparedtothepurchaseandoperationalcostsofmostvehicles.

SeveralofthePAMsystemsarerecordingonly,whereasotherspurporttoofferreal-timedetection.Perhaps

themostsuccinctlyusefulcommentcameinareplyfromCornellUniversityregardingtheautomaticdetection

capabilitiesoftheirAutoDetectionbuoys:“…anyspeciesforwhichatrustedalgorithmexists”.Asdiscussedin

section 9.5.2, an increasingly wide array of automatic detection algorithms is appearing in the literature,

however,mostofthesearetunedforveryspecificanalysisofhistoricaldataandareunlikelytoperformwellin

many situations. PAM systems generally require a human operator to verify detection data and without a

considerable quantity of data available for human verification the chances of false alarms from automatic

detectorscanbehigh.

Forlongtermmonitoring,themostpracticalsolutionsareprobablythelowestpowersystemavailablewhich

canrecordasmuchdataaspossibleandperhapsperformaminimumofdatareduction.Keytoanymission

planningisdefiningtherequiredupperfrequencylimitofthesystemsincethiswillcontrolthemissionduration

availablewithagivendatastoragecapacity.Eitherallrawdatacanbestored,orsimplealgorithms(employed

63


DATE:July2016


torunatbothhighefficiencyandhighfalsealarmrate)canbeusedtoreducethequantityofstoreddata,which

isparticularlyusefulwhenworkingwithhighfrequencyspecies.

Formitigationmonitoring,thenmissiondurationmaybeoflessimportanceandprioritymaybegiventodata

relay,on-boardprocessingandeaseofconnectivity,whichisprovidedbythehigherpowersystems.Underwater

vehiclesystemsclearlyhaveadisadvantageformitigationmonitoringgiventhetimedelaysincurredwhilethey

returntothesurfacetotransmitdata.However, forsomemitigationscenarios,wherespeciesareknownto

moveslowly,theremaystillbeauseforthesevehiclesinmappingoutareasunlikelytocontainmarinemammals

so thatoperations canminimise the likelihoodof incurringa shut-downwhilst real-timemonitoringaround

thoseoperations.

None of the systems reviewed had any significant operational requirements. However, for real-time

interpretationof incomingdata,trainingandexpertisewouldberequiredofasimilar levelas isrequiredfor

PAMoperatorsonseismicvessels.

Physicaldimensionsofthesystemsvariedfrom9.5x4x2.5cmforthesmallestsystemto56x13x11cmfor

the largest.Weights ranged from120g to1.5 kg.Power requirementwasalsohighly variable,with several

systemsusing lessthan100mW,the lowestbeing35mWforthesinglechannelSoundTrapHF.Thehighest

powersystemsusedbetween2and4Wofelectricalpower.

Mostsystemshavetheabilitytostorerecordeddata,mostlyusinglosslesscompressionalgorithmssuchasFLAC.

However,evenwithterabytesofinternalstorage,thisdoesnotovercomethefundamentallimitationsofstoring

dataacquiredathighsamplerateoutlinedinsection9.5.However,asstoragecapacitycontinuestoincreaseit

islikelythatfullrecordingsofevenhighfrequencysoundwillbecomeincreasinglyaccessibleinthecomingyears.

RawPAMdatacomesinatarateofamegabytepersecondforhighfrequencyspecies,100kilobytespersecond

formid-frequencysoundssuchasspermwhaleclicksanddolphinwhistlesandafewkilobytespersecondfor

low frequency sounds, such as baleen whale calls. Some of the data relay systems using either analogue

transmission or Wi-Fi based digital technology are capable of transmitting that data volume over several

kilometresusingfreespectrum(i.e.nodatarelaycostsbeyondequipmentpurchase).Powerconsumptionfor

thesedatarelaysystemssbetween10and30W,whichismanytimeshigherthanthepowerconsumptionof

themonitoringsystemsthemselvesandwouldimpactonthelifetimeofmostvehicledeployments.However,

formitigationmonitoring,wheredeploymentsdurationsofhourstodaysmaybeacceptable,thismaynotbea

problem.

Thehighestpowersatellitebasedsystemscanrelaydataat128kbpsor16kBytespersecond.Thiswouldbe

insufficient fortransmissionofanythingbut lowfrequencyrawdata.Thesesystemsarealsophysically large

(tensofcentimetresineachdimension),highpower(100–150W)andcostthousandsofdollarspermonthto

operate.Smaller,lowpowersatellitesystems,suchastheIridiumL-Band,areavailablewhicharemoresuitable

64


DATE:July2016


formountingonautonomousplatformsandonlyconsume2.5Wpower.However,thelowpowersystemshave

anevenmorerestrictedbandwidthof2,400bps(300bytespersecond,orarawacousticbandwidthof150Hz).

Theonlysystemsweareawareofwhichhaveachievedsufficientlyreliablealgorithmssuchthatadequatedata

canbesenttoshorevialowbandwidthsatellitelinksaretheunderwatergliderbasedDMON(Baumgartneret

al.,2013)andtheCornellAutoDetectionbuoys.However,bothofthesesystemsonlyachievethatlevelofdata

reduction for certain species of baleen whale. None of the systems reviewed have to date sent adequate

informationviasatellite linksforhighfrequencyodontocetesoundsforreliablereal-timedetectionwith low

falsealarmrates.SeveralofthesystemsofferserialandEthernetbasedconnectivityandtheSAInstrumentation

LtdDecimusandtheSeichesystemcanbothbeusedwithwirelessmodemsystemstosendenoughdatainreal-

timeformitigationmonitoringofawiderangeofspecies.

To summarise: for real-time mitigation monitoring, high frequency real-time data can be sent over short

distances (up to a fewkilometres)using free spectrum radio links.Where local radio links arenotpossible,

accurate identification of marine mammals is currently only practical for some baleen whale species. For

populationmonitoring,whereit isunlikelythatradiolinkscanbeused,rawdata,orpartiallyprocesseddata

shouldbestoredforanalysispostrecovery,withonlysummaryandstatusdatabeingsentthroughsatellitelinks.

8.2.4.2 AAM

CurrentlythereareselectedAAMsensorsandAUVgliderplatformsthathavealreadybeenintegrated.These

include:theImagenexES853andNortekADCPsensorintoboththeSlocumG2gliderandtheSeaglider(ogive),

aswellastheImagenexES853sensorintotheSlocumG2hybridandtheSontekADPsensorintotheSCRIPPS

Spray.ThecomparisonmatricesshowedthattheImagenexES853,Kongsberg/SimradWBTmini,SontekADP,

NortekADCPandtheVemcomodularVR2Ccurrentlyhavethecapacitytobeintegratedintoanyofthelisted

AUVglidersystems.TheBiosonicsDT-X-Sub,miniWBATandtheASLAZFPcouldpotentiallybeintegratedinto

anyofthelistedAUVglidersystemswhilstthesesystemsandtheSimradWBATcouldbeintegratedintolarger

AUV.ABiosonicsmultifrequencyinstrumenthasbeenintegratedintoaWaveglider(Greene,2014).

Wheregiven,pricerangesforAAMsystemswereallinthe$2,000to$100,000pricerange.However,thisisthe

purchase price for the sensor and therewill likely be additional costs associatedwithmounting any of the

systemsintospecificvehicles.Ofthesensorsthatareavailabletorent,thepriceforathree-monthrentalranged

between$5,000and$20,000.

ThedatacollectedbytheAAMsystemisraw,unprocesseddata,withtheexceptionoftheBiosonicsDT-X-Sub

whichhason-boardprocessorsthatproducesdatasummarieswhichcanbeviewedassimplifiedechogramsand

isabletoprovidealertstriggeredbyacousticevents.MostoftheAAMsystemsareableto identify fishand

zooplanktonfromtherawdatacollected,aswellasprovideabearinganddirectrangeestimatetotheanimal

fromasingledevice.Conversely,theVemcomodularVR2CandVMTcanonlybeusedtodetectspecieswhich

arelargeenoughtobetaggedwithanacoustictagandcannotprovideeitherabearinganddirectrangeestimate

65


DATE:July2016


totheanimalfromasingledevice.Real-timedatatransmissionispartlypossibleforallAAMdevices(apartfrom

theVemcoVMT)whereapingcanbetransmittedinrealtime.Howeversinglepingshavelimitedvalueexcept

toindicatetheinstrumentisworking,analysesaretypicallyundertakenonmultiplepings.

Whereprovided,thelevelofautomationfortheAAMsensorswascategorisedasself-loggingwithafixedping

rateorprogrammedintervals.

TherawdatavolumevariesgreatlybetweenAAMsensors,dependingontheircomplexity.Forexample,theraw

datavolumefortheImagenexES853is256bytesperping,comparedtothe32bytesperpingfortheSontek

ADPand5,6124to1,159bytesperpingfortheASLAZFP(dependingonthenumberoffrequencies,binsizeand

range).MoreinformationonthedataprocessingandtransferinformationwasprovidedfortheBiosonicsDT-X-

sub,whichhasarawdatavolumeof30MB/sec,theabilitytoprovidelowbandwidthsummaryreportingasa

methodofreal-timedatareduction,apowerusageof30wattsandarequiredexternalpowerfeedof12to24

voltsDC.

8.2.4.3 Animal-bornetags

ItispossibletointegratetagreceiversintoanAUVorASVtoprovidereal-time/near-real-time(dependingon

glidercommunicationstrategy)information(e.g.Waveglider,SlocumorREMUS).Vemcotaglocatorshavebeen

integratedintoaSlocumglider(Haulseeetal.,2015),REMUSAUV(Eileretal.,2013)andaWaveglider7.PAM

systemsasevaluatedinsection8.2.4.1canpotentiallybeusedtodetectthesignatureofacoustictags(Sparling

etal.,2016).Wedidnotconsiderthesesystemsfurtherasithasaminorroleinmonitoringwithregardstonoise

impact.

8.3 Discussion

8.3.1 How to use the evaluation results to find the appropriate autonomous platform / sensor

combination

Thisreviewpresentsa largevarietyofautonomousvehiclesandsensors thataresuitable formarineanimal

monitoringduringE&Pactivities.Therequirementsformarineanimalmonitoringwerethedriversbehindthe

informationthatwasgatheredforeachofthesesystems.Thecompiledcomparisonandevaluationmatrices

(seesections11.3and11.4aswellasTable2,Table3andTable4)helptoselecttheappropriateplatform/sensor

combination, and the followingdiscussionpointsoutwhat capabilities are required from theplatformsand

sensorsinvariousmonitoringsituations.Anyonewishingtousetheinformationprovidedinthisreviewshould

definetheobjectivesandrequirementsofthemonitoringfirst.Forexample(thisisnotacomprehensivelistof

questions):

7 http://oceantrackingnetwork.org/otn-tests-new-wave-glider-technology/#prettyPhoto

66


DATE:July2016


• Whichanimalspeciesorspeciesgroupshallbemonitored?

• Isitmitigationmonitoring/populationmonitoring/fine-scalebehaviouralmonitoring?

• IfduringE&Pactivity,whatkindofactivity?

• Isthemonitoringareanearshoreoroffshore?

• Whatisthesizeofthemonitoringarea?

Speciesorspeciesgroupwilldeterminethekindofsensorthatcanbeusedformonitoring.Anextensivereview

on which sensor is best suitable for which type of animal in which environmental conditions (including

geographicallocation)wasrecentlyconductedforIOGPandsummarisedinVerfussetal.,(2016)andistherefore

kepttoaminimuminthecurrentreview.

AAMcandetect anythingwith a largeenough target strength tobe capturedby its receivingelements, i.e.

enoughenergyfromtheemittedsonarpulsesneedstobereflectedbytheanimals’bodytobeimagedonthe

sonar screen.This iswhyany large “body”, fromzooplanktonpatchesand fishesup to largewhales canbe

detected(Table2).Classificationoftheanimalspeciesorspeciesgroup is,however,complexwithAAMand

reliesonknowledgeoftheanimal’sspecifictargetstrengthatdifferentfrequenciesandswimmingbehaviour.

Specificalgorithmsfordetectingandclassifyingmarineanimalssuchasseals,porpoise,dolphinsandsharkswith

theTritechSeaTecsystemhavebeendevelopedorarestillindevelopment(asdiscussedinsection7.3.3).These

mayhowever,onlyworkinthecontextforwhichtheyweredevelopedandtherefore,mayneedtobetested

forothersituationsorenvironments.

PAMonlydetectsvocalisinganimalsandismainlydevelopedforwhaleanddolphindetection.Whichcetacean

speciescanbemonitoreddependsmainlyonthefrequencyrangeofthesystem,thekindofdetectorusedand

whethertonalcallsorecholocationclickscanbecaptured.Speciesidentificationisrestrictedtoanimalswith

verydistinctivecallsorclicksthatcanbedistinguishedfromotherspeciesanddependsonthesoftwareused

with the systemand the entailed classification algorithms.Detection algorithms canoftenbeused for data

reduction,filteringthemostlikelydetectionsoutofthehugeamountofdataretrieved,butgenerallyrequire

visualconfirmation,especiallyformitigationmonitoring,wherefalsealarmsmaycausedelaysinoperations.

WhilePAMandAAMaresensorsthatareusedwithAUV/ASVandarethereforespecialisedtounderwaterlife,

video sensors are coupledwithUAS andworkwell for animals surfacing frequently. Thismeans that video

sensorsmaybeinefficientfordeepandlongdivingspecies,atleastformitigationpurposeswherethedetection

probability would need to be high (see Verfuss et al., 2016). With appropriately trained human observers

monitoringthevideostream,classificationofanimalswiththiskindofsensorispossible.Inordertoreducethe

hugeamountofdata collectedbyvideo sensors, algorithmscanbeused topresent likelydetections to the

observer,thusdecreasingthepotentialworkloadandincreasingobserverefficiency.

67


DATE:July2016


Thekindofmonitoringdetermines theplatform requirements. Thesearedetailed in section9.1.Mitigation

monitoring, for example, relies on continuous real-time detection capabilities in order to enable a timely

decisiononwhethermitigationmeasuresare tobe takenornot (section9.1.1). Forpopulationmonitoring,

sendingdatainreal-timeisnotrequiredbutmaybeadvantageous(section9.1.2).Focal-followsmayormaynot

requirereal-timeoption,dependingonwhetherthevehiclesarebeingpilotedmanuallyorusingdetect-and-

trackcapabilities(section9.1.3).

Table2givesanevaluationofthosesensorsincludedinthisreviewwithreal-timedetectioncapabilities,and

whichspeciesgrouptheyareabletodetectorevenclassify.

Formitigationmonitoring,wherethecontinuousreal-timemonitoringisrequired,thesensoraloneisnotthe

determiningfactor.Theplatformneedstobecontinuouslyincontactwiththedatarelaysystem.Anysensor

mountedonanunderwaterautonomousvehiclecanonlysenddatabackwhenattheseasurfaceandincontact

withtherelaysystem.Therefore,onlysurfaceoraerialsystemswouldbefeasibleinmonitoringandmitigation

instanceswhencontinuousreal-timedetectionisrequired.

68


DATE:July2016


Table2.Evaluationmatrixofdetection/classificationcapabilitiesofsensors(includedinthisreview)capableofreal-timedetection.Withthehelpofthesesensors,oneiscapableofdetecting(Detect)ornotcapableofdetecting(No)marineanimals.Someofthesensorsdeliverdatawithwhichmarineanimalscanbeclassified(tosomeextent)tospeciesorspeciesgroups(Classify).

Systemclass Systemname Realtime

Zooplankton

/Fish

Whales

Dolphins/

Porpoises

Seals

Turtles

Largefish

AAM

Aquadopp Partly

Detect

ADP Partly

AZFP Partly

DT-XSUB Partly

ES853 Partly

Gemini720i YesDetect Classify Detect Classify

Gemini720is Yes

ModularVR2C+tag Yes Classify

WBAT PartlyClassify Detect

WBTmini Partly

PAM

C-POD-F Yes

No

NoClassify

No

Cornell/AutoBuoys Yes

Classify

Decimus Yes

DMONYes

(Whalesonly)Detect

SDA14 Yes

Classify

SDA416 Yes

Seichereal-time

transmissionsystemYes

WISPR Partly

VIDEO

CM100 Yes

Detect Classify

CM202 Yes

DualImager Yes

EO900 Yes

OTUSU135HIGHDEF Yes

OTUS-L205HIGHDEF Yes

TASE310 Yes

TASE400HD Yes

69


DATE:July2016


ThekindofE&Pactivitywouldfurthermoredeterminethechoiceofplatform,e.g.anyAUV/ASVaroundseismic

surveyvesselswouldcreateasignificantsafetyrisk(seesection9.3.5)andwouldthereforenotbesuitablefor

mitigationmonitoringintheseareas(Table4).Ontheotherhand,deploymentofanAUV/ASVinthevicinityof

an explosive removal operation for a period of monitoring prior to detonation might be relatively

straightforward.Thelocationofthemonitoringareawilldeterminewhichdatarelaysystemtobeusedaswill

themonitoringkind.Mitigationmonitoringoftenrequiresthetransmissionofalargeamountofdatatoahuman

observerforthepurposeofverifyingdetections.Therefore,ahighbandwidthisrequired,suchasisprovidedby

awireless system. For locations furtheroffshorehowever,wireless systemsareoftenunavailable (Table3).

Satellite systems, by contrast, have an unlimited communication range thatwould assist the data transfer,

however,thedisadvantageisthelowbandwidththatsatellitesystemsemployingeneral.Theinstallationofa

wirelesssystemonavesselaccompanyingtheautonomousvehiclewouldbeanothersolutionforshortrange,

highbandwidthcommunicationoffshore(formoreinformation,seesection9.5).

Table3.Propertiesgenerallyapplyingforautonomousvehiclesforthedatarelaysystemssatellitelinksandwirelesssystems

Property Satellitelinks Wirelesssystem

Bandwidth Low High

Communication

rangeUnlimited Short

Thesizeoftheareamonitoredwilldeterminethetechnicalrequirementsofaplatformclass.Largeareaswill

demandlargerandmorerobustvehicleswhilesmallerandmoreagilesystemsmaybemoreappropriatefor

smallermonitoringareas.Forexample,inthecontextofoffshoreseismicsurveys:smalllow-rangeUASmaybe

usedtomonitormarinespecieswithintheexclusionorhigh-riskzone;whilelongrangeUASmaybeusedto

obtaininformationbeforeandafterthearrivalofseismicvessels,thusgatheringinformationonanimaldensity

anddistributioninthespecificregiontargetedforoperations(baselinemonitoring)(Wattsetal.,2010).

Theabovetextisanexampleforhowtoapproachthedatagatheredinthisreview.Thesubsequentsectionswill

discussandexplain furtherdetails tobe considered.Thisoverviewofautonomousvehicleswill support the

selection of the appropriate combination of systems for data acquisition and processing, considering the

differentqualitiesofplatformsandsensorsthatareapplicabletodifferenttypesofsurveys.

Ashighlightedabove,thediversityofapplicationsforautonomousvehiclesmeansthatspecificsurveyscenarios

arerequiredinordertoprovidespecificrecommendationsonwhichplatformandsensorcombinationmightbe

best to use in a specific scenario. Given the large number of target species, selection of the appropriate

sensor/platform type is possible rather than selecting specific individual systems. The same is true for the

differingmonitoringtypes,combinedwiththechoiceofdatarelaysystem.Thetechnicalandoperationaldetails

will need to be tailored to the specific needs of the E&P project, e.g. the area of interest, the platform an

70


DATE:July2016


autonomousvehiclewillstartandfinishoperatingfrom,therequiredsurveylength,projectbudgetandmore.

Herewegiveanoverviewofwhichplatformtypesarebestforwhichmonitoringtypebycomparingeachclass

ofplatformsandhighlightingtheirparticularstrengthsrelatingtothespecificmonitoringtypes.Theresultsare

summarisedinTable4.ExamplesofspecificE&Pprojectoroperationsneedswillbegiventonarrowthechoice

ofplatformsbestsuitedtoaspecificapplication.Onecouldthenchoosetheappropriateplatformandsensor

fromtheshort listsgiveninsection11.2,basedonthetechnicaldetailsandoperationalaspectsgiveninthe

comparisonmatricesinsection11.1thatsuitthemonitoringapplicationconsidered.Theevaluationmatricesin

section11.4,showingwhichsensormayfitintowhichplatform,thenhelpstofindtheappropriatecombination

ofthosesystems.

Itis,however,importanttonotethattheuseofautonomousvehiclesforanimalsurveyingwillencountermany

of the same survey design, data collection and analysis issues currently experienced when using standard

mannedsurveyvehiclesorstaticPAM.Forexample,itshouldbeshownthatanyautomaticdetectorsusedare

abletoperformadequatelygiventhesurveyingconditionse.g.inpoorweatherconditions(forbothvisualand

acoustic data) or in the presence of local industrial or ambient noise sources (acoustic data). Furthermore,

suitablehardwareshouldbeselectedifestimationofdetectionprobabilities,orothermultipliers,isrequired.

Table4.Evaluationmatrixofmonitoringcapabilitiesofsensor/platformtypes.Giventheymeetthecriteriaasoutlinedinsection8,thespecificsensor/platformtypeslistedareingeneralwellsuited(�),suited(x)ornotapplicable(-)forthecertainmonitoringtypes(exceptionsmaybepossible).Themonitoringtypesarefurthermoredividedintomitigationmonitoringinareaseitherclearfromorbusywithotheroperationalgearortraffic,short-term(hours,days)orlong-term(weeks,months)populationmonitoringandfocal-followsconductedwithstaticormobilesystems.L-t-a=Lighter-than-airaircraft.

Monitoringtype

Mitigation Population Focal-follow

Sensor Vehicle Clear Busy Short-Term Long-Term Static Mobile

Electro-optical Poweredaircraft � � � - x �

Motorisedgliders - - - - - -

Kites � � x - x x

L-t-aaircraft � � x - x x

PAM

PoweredAUV � - � - � x/-

ASV � - � - � x/-

Self-poweredAUV � - � � � -

ASV � - � � � -

Drifter - - � � - -

AAM

PoweredAUV � - � � � �

ASV � - � � � �

Self-poweredAUV � - � � � �

ASV � - � � � �

Drifter - - � � - -

71


DATE:July2016


8.3.2 Strengthandweaknessesofautonomousvehicletypes

8.3.2.1 UAS

UASgenerallyoperateatalowerspeedthanmannedaircraftandwithasmallerfieldofview.UASmay,however,

compensateforthisbysurveyingforlongerperiodsoftimeandreachingareasinaccessibletomannedaircraft.

Weatherconditions,suchasseastate,havebeenreportedtoaffectsightingrateofmedium-sizedmammals

lesswhenusingUASimagingcomparedtohumanobservers,whichisoneofthemainlimitingfactorsknownto

decreasesightingratesinmannedsurveys(Hodgson,2013).ThevariabilityinUASsizescurrentlyavailableisalso

agreatadvantageforoffshoremonitoring,sincesmalleraircraftallowformonitoringofsmallerareasatalow

cost (e.g. exclusion zones during seismic operations), and larger aircraft can cover larger areas and provide

valuable informationbothbefore,during, andafteroffshoreoperations. The longoperational rangeofUAS

mean that they are also quite suitable formitigationmonitoring of the area ahead of a seismic vessel for

detectingmarinemammalswithinanexclusionzonearoundtheanticipatedstartlocationofthesoundsource,

assuggestedbyVerfussetal.,(2016).

Inadditiontothepotentialuseformonitoringmarineorganismsusingvisualsystems,UASmayalsobeusedto

drop various types of autonomousmonitoring systems at pre-determined locations. In the case of marine

mammalmitigationinrelationtoseismicsurveys,itmaybepossibletodrophydrophoneswithabuilt-inradio

link(e.g.sonobuoys8)toacousticallymonitorareaspriortothearrivaloftheseismicexplorationvessel.This

would allow the operators to optimise operations planning given this “early warning” of potential marine

mammalpresence.

Hence, it ispossible to coverawide rangeof studiesusing the same technology.However, thereare some

shortcomingstotheuseofUAS,suchastherestrictedfieldofviewofanimageandtheimageresolutionrequired

tocompensateforthisathigheraltitudes,andthetimetoanalysethedataafterthesurvey.Forthislastfactor,

real-timedatatransmissionand/orautomateddetectionofanimalswouldhelptoreducepost-survey image

analysistime.

Theadvantagesofusingunmannedsystemsnotonlycoverhealthandsafetyissuesbutalsothequalityofthe

datathatcanbeacquired.Datafromdigitalsurveysovercometheissuesrelatedtoobserverbiasandaccuracy

oftheinformationconcerningflightoperationdata(altitude,speed,fieldofviewaccuracy).AcquiringtheGPS

locationofeveryimageprovidesgreateraccuracyinthelocationofadetectionthanformannedsurveyswhere

8 See http://www.uasvision.com/2015/07/22/ultra-electronics-studies-uav-drop-options-for-sonobuoys, https://www.google.com/patents/US3986159 (last accessed 03.02.2016)

72


DATE:July2016


observersreportthesightings.UsingtheGPSlocationofanimage,itispossibletoobtainthecoordinatesfor

theimagecorners,andthusofeachanimalsightedwithinit(softwareVadar9).

ThesuccessrateintheuseofUASishighlydependentonthemainaimsforitsuse.Circumstancesthatdemand

long range operations with higher sensor quality will thus be pricier than in circumstances where focal

monitoringofsingleindividualsorgroupsofanimalsisneeded,whichmayprovidevaluableresultswithsmaller

andlesscostlyplatforms.Althoughthereareseveralhundredsofplatformsandsensorsonthemarket,andnew

systemscontinuouslybeingdeveloped,itisobviousthatmanysystemsproducegoodqualitydataregardlessof

theplatformused.Thecombinationbetweenplatformandsensorshouldbechosenbasedontheoperational

needsandthecapacityfortheplatformtocarryaspecificsensorfortheamountoftimeordistancerequired.

Dependingonthepayload,alow-costcamerawithhighorlowstabilityandqualitycanbemountedintothe

system,providingvaluableinformationaboutanimaldistribution,abundance,andbehaviour.Balloonsandkites

arecost-savingalternativesforsomefixedandrotarywingsystems(ThammandJudex,2006;Altanetal.,2004;

Fotinopoulos, 2004; Grenzdörffer et al., 2008; Scheritz et al., 2008), though they rely heavily on weather

conditionsandaredifficulttocontrolwhenattemptingtofollowapredefinedtrack.Intermsofdataanalysis,it

isoftensufficienttohaveaquickoverviewofimagesorvideosacquiredusinglow-costsensors,whereasfor

accuratemeasurementswithmoreprecisemethods(e.g.differentialGPS),datarequireamoredetailedreview

and evaluation. In recent investigations, the trend seems to be directed to fast processing such as online

triangulationanddirectgeo-referencing,whichmayfacilitateimageryanalysisandreducepersonnelcosts.In

addition,thesensorsandsystemsaregettingsmaller,atafractionoftheprice,andoperatingwithopensource

platformsthatarecontinuouslybeingdevelopedinphotogrammetry.Dataintegrationofvarioussensorsand

highprecisionandresolutionofspecificapplicationsandnearreal-timeapplicationshasalsobeenunderfocus

recently,withcontinuousprogress.

ThisstudymakesthefollowingshortlistofrequirementsforaUASmonitoringsystem:Stabilisedhigh-resolution

videofordetection,highresolutionimagesforidentification,real-timegeocodingofimagedataandacomputer

systemforefficientanalysisofrecordeddataandlocalisationofanimals.Ifthedataaretobetransmittedtothe

groundahighbandwidthdatalinkisalsorequired.

8.3.2.2 AUV/ASV

ThereisabroadrangeofAUVandASVmodelsandsensorsthatwillmeetmostneeds.Theplatformshouldbe

chosen based on the size, endurance, payload and depth capabilities required for each application. Larger

poweredAUVaregenerallymoresuitablefordeepwater,openseatasks,thansmallpoweredAUVwhichare

moresuitedtoshortermissions,wherethereducedcostisamustandthesmallerpayloadisnotalimitation.

9 www.cyclopstracker.com

73


DATE:July2016


UnderwaterbuoyancyglidersoffersimilarcapabilitiesforsensorcombinationsaspoweredAUV.However,they

havereducedpayloadcapabilities,reducedmanoeuvrabilityinfastcurrentsandoperateatslowspeeds.Onthe

positiveside,theyhavelongmissiondurations,independencefromamothershipandtheabilitytoprofilethe

verticalrangeoftheocean.Theintegrationofbothactiveandpassivesensorsintounderwaterbuoyancygliders

isalsorelativelywelladvancedinthescientificcommunity.

Self-poweredcraftsuchastheWavegliderofferlargelysimilarcapabilitiesaspoweredASVbuttherearesome

maindifferences.Reducedpayloadcapabilities in self-poweredcraftmaybea limiting factorbut the longer

missiondurationandgreater independence(intermsofminimisedmechanicalmaintenance)ofsimplerself-

poweredASVhassignificantappeal(particularlyforlongertermenvironmentalmonitoring).

8.3.3 Platformtypesformitigationmonitoring

MitigationmonitoringgenerallytakesplaceinthepresenceofsomekindofE&Pactivitymeaningthatitishighly

likelythatanumberofvesselsareoperatinginthearea.Thesevesselsmayhavecompetenttechnicalpersonnel

on board with the ability to deploy, service and operate autonomous platforms. However, in many

circumstances,vesseloperationsarealreadyworkingclosetothelimitofwhatispracticallyfeasibleintermsof

equipmentdeployment,personnelarealreadyassignedtochallengingandall-consumingtasksandtheremay

notberoomonboardeitherforadditionalequipmentorforadditionalpersonneltooperateit.

Unmannedvehicleshavesofarnotbeenusedformitigationmonitoringbutbringpotentialintothisfield.Some

systemsmaywellbeplacedfortheuseofmitigationmonitoringastheyallowfor(near)real-timedetection,

whichisnecessaryfortimelydecisionmaking.Thesesystemswouldneedtooperateforlongenoughperiods

and cover wide enough ranges to meet the temporal and spatial requirements of the various guidelines.

Synchronisedoperationofasmallorlargerfleetofvehicleshasthepotentialtoobtainawiderfieldofviewto

cover larger areas than single vehicles. The concurrentuseofdifferentmitigationmonitoringmethods, e.g.

different autonomous sensor/platform types combined with monitoring methods installed on a manned

platform(e.g.MMO/PSO,thermal-IR)will increasetheprobabilityofanimaldetection,whichisanimportant

factorinmitigationmonitoring.

Mitigationmonitoringisusuallyconductedclosetothemitigationzoneasternoftheseismicvessel–anarea

wherenovessel,letaloneanautonomousone,cansafelyoperate.Othersupportvesselsdooperateaheadand

tothesidesofseismicvessels,but,asoutlinedinsection9.2.1,therewouldbesevereproblemsindeploying

additionalautonomousseabasedvehiclesintheseareas.ThisiswhyAUVandASVarenotconsideredsuitable

formitigationmonitoringinoperationallybusyareas(Table4).Aerialsystemshowever,maybesuitedifthey

meettherequirementsoutlinedinsection9.1.1.

Indynamicoperations(e.g.aseismicvessel)thereisgreatpotentialforautonomoussystem,especiallyUAS,to

scopeahead,tooperateasanalertsystemtomitigationmonitoringmethodsonthesourcevessel. Inmany

operations, the unmanned systemswould be at a stand-off distance from the source, whichmay (but not

74


DATE:July2016


necessarily)limittheabilitytoestablishtheanimal’spositonrelativetothemonitoringzoneorsoundsource,

respectively.However,onebigadvantageofautonomoussystems is thattheygivetheopportunitytocover

great distances and large areas, especially when using multiple systems that coordinate their movements

throughaco-operativecommunicationsystem.However,asdiscussedabove,therearesignificantlogisticaland

safetyconstraintswhichwouldneedtobeovercomebeforeseabasedautonomousvehiclescouldbeoperated

inthevicinityofaseismicsurvey.

As discussed in section 9.1.1.1, the localisation of the target animal would be a great advantage. The

experimentalautonomousPAMsystemshavetendedtouseonlyoneorasmallnumberofhydrophonesclose

together. In most cases, these can only provide animal presence; some may be able to provide bearing

informationtotheanimal,butnonecouldprovideanimalrange.Theonlyexampleofanexceptiontothisisthe

simultaneousdeploymentofthreesubmarineglidersofftheUSEastCoast(Fuciletal.,2006),wheretimeof

arrivaldifferenceofthesignalatthethreevehicleswasusedtoestimateanimallocation.Whilethistechnique

couldbeimplementedusingmultiplesurfacevehicles,ifthedeploymentofonevehicleinthevicinityofaseismic

survey isproblematic,thendeployingthree(ormore)vehicleswouldclearlyrequirecarefulconsideration.A

vehiclecapableoftowingmorethanonearrayorequippedwithbeamformingcapabilityorvectorsensorscould

alsoresolveleft-rightambiguitiesandlocalizeinnearrealtime.

Ifitwerepossibletodeployvehiclesclosetoaseismicsurveyvessel,automaticdetectorscanbeusedwithsome

species, but false alarm rates are rarely low enough for action to be taken without some verification, so

significantamountsofdataneedtoberelayedbacktoanoperator.Therelayofsignificantquantitiesofdata

backtothevesselisaproblem,butnotaninsurmountableone,sinceanumberofwirelessradiosystemsexist

whichcansendwidebandwidthdataovershortdistances.

Inoperationssuchasunexplodedordnancedevice(UXO)removalorwell-headdecommissioning,wherethe

sourceisinafixedpositionbutanMMO/PSOordetectiontechnologycannotbeplacedcloseenoughtothe

source to provide adequatemonitoring, autonomous vehicles would provide a safe and effective solution,

particularly,astheydonotplacehumanswithinadangerousworkingarea.

TrialswithPAMonASVC-WorkerandC-Enduroshowedpromisingresultswithsuccessfulreal-timedetections.

AnappealingideaisfortheASVtoscopeaheadoftheseismicvesselformarinemammals.Thisoffersthebenefit

of operating in lower noise environment (away from the vessel), and would enhance ability to detect low

frequencybaleenwhalevocalisations,whicharehardtodetectwithPAMinstalledfromaseismicvessel(Verfuss

etal.,2016).However,thiswouldnotcoverthemonitoringzone(operatinginsteadasanearlywarningsystem).

So it would not replace existing methods, but would be complimentary, or could be used for assessing a

monitoringzonelocatedaroundtheanticipatedstartofsoundsourceoperation(e.g.duringseismicsurveys)as

suggestedbyVerfussetal.(2016).ThistaskcouldalsobeachievedusingUAS.

75


DATE:July2016


8.3.4 Platformtypesforpopulationmonitoringandfocal-follows

8.3.4.1 UAS

Ofthethreeclassesofaerialplatformsconsideredinthisreview,poweredaircrafthavemanyofthecapabilities

requiredforaerialsurveysofmarineanimals.Kitessharemanyofthesameattributesaspoweredaircraftand

theyareeasiertotransportand,insomecases,deploythanfixedwingaircraft.However,theyarelessrobust

tobadweather,whichisamajordisadvantageforwildlifemonitoring,especiallyduringoffshoreoperations(see

Table4forthisandanyfurtherdiscussion).Sensordatafromtheseplatformscanbegeo-referencedusingdata

from a GPS and flight data recorder,which greatly improves the accuracy ofmeasurements related to the

individualanimal’spositionwhensighted.

Lighter-than-airplatformsalsohavecapabilitiestoperformmarineanimalsurveysbut,duetotheirrequirement

foratether,severalwouldeitherhavetobemooredtostaticbuoystocreateaseriesofmonitoringpoints,or

attachedtoamovingvessel (whichcould, intheory,alsobeautonomous)toconducta linetransectsurvey.

However,bothofthesescenarios(mooredandtethered)arepotentiallymorelogisticallycomplexthanusing

fixedwingUASand,therefore,poweredaircraftappeartobetheoptimalcandidateplatformforaerialtransect-

basedsurveysusingautonomousvehicles.

In terms of individual focal follows, however, lighter-than-air aircraft should be considered as a potential

candidateplatform,thoughtheirabilitytofollowanimalswillberestrictedtothemanoeuvrabilityofthesupport

vesseltowhichtheyaretethered.Poweredaircraftcanbepilotedtofollowanimals,rotarywingvehicleshave

performedwellinexistingstudies(Durbanetal.,2015),thoughfixed-wingaircraftmaystruggletohoverabove

stationary,orslowmoving,focalanimals.Therefore,poweredaircraftandlighter-than-airaircraftcouldbegood

candidateplatformsforindividual-basedmonitoring,dependingontheexactgoalsofthemonitoring.

Inconclusion,thestudiesmentionedpreviouslyandsystemsevaluatedforthisreviewhighlightthepotentialof

UASinmarinewildlifemonitoring.Koskietal.,(2009a)andHodgsonetal.,(2013)supporttheuseofunmanned

aircraftinmarinemonitoringanditspotentialtoreplacecurrentmethodsemployinghumanobserversposted

onthesourcevesselor inanaircraft.However,asmentioned inKoskietal., (2009a), thiscanonlyreallybe

achievedwhensurveyinglargeindividualsorlargegroupsofsmallanimalsifthesearchareaissmall.Forlarger

areasanddetectionofsmalleranimals,higherresolutionvideoisrequiredifoneistocomparebothmanned

andunmannedaerialdetection.Therefore,thenextstepsinestablishingtheefficacyofreplacingmannedaerial

surveyswithUASistoconductadirectcomparisonanddeterminewhethertheproportionofanimalsavailable

fordetectionfromtheairisequivalentforbothmethods.Thiswouldneedtobeconductedunderavarietyof

conditions,rangingfromArcticsummer/wintertotemperateandtropicalregions,whereboththedetectability

andenduranceofUASisexpectedtovary.Additionally,testingatdifferentlocationscouldalsoincludearange

ofdifferentspecies,whichwouldassistindeterminingthelimitationsofspeciesidentification,particularlyin

regionswithlong-divingandevasivespecies.

76


DATE:July2016


8.3.4.2 AUV/ASV

ThefiveclassesofAUV/ASVreviewedinthisreportareallcapableofcompletingmarineanimalsurveysof

various kinds (see Table 4 for this and any further discussion). Some classes of platform share similar

characteristics, though there are also clear different strengths andweaknesses between classes relating to

marineanimalsurveying.

Thepotentialtoincreasespatialandtemporalsurveycoverageisamajorbenefitofusingautonomousvehicles.

Tothatend,thefactthatpoweredsurfacecraftandpropellerdrivenunderwatercrafthavesurveyduration

times on the order of hours, rather than the months that unpowered vehicles can be deployed for, is a

disadvantage for long-term wildlife surveying (though a notable exception was the ASV C-Enduro, with a

maximumsurveydurationof90days).However, theseplatformsmaybemore suitable for individual focal-

follows,especiallythosewithincreasedmanoeuvrability.

The autonomous underwater buoyancy gliders and the self-powered surface vehicles generally havemuch

longersurveydurations,whichmakethemveryattractiveforlong-termwildlifemonitoring.Furthermore,the

abilityofunderwatergliderstoverticallyprofilethewatercolumn(whichmayalsoaiddetectionofdifferent

marine animals that occur at varying depths) is a further advantage of these vehicles. Reduced payload

capabilitiesmaybea limiting factorbut the longermissiondurationandgreater independence (in termsof

minimisedmechanicalmaintenance)ofsimplerself-poweredASVhassignificantappeal,particularlyforlonger

termenvironmentalmonitoring.

Animportantadvantageofthesmaller, lowpoweredautonomousvehiclesistheirlownoisefloorwhichcan

provideatleastsomelevelofconfidencewhencomparingsurveyresultsfromdifferentregionssincea)masking

ofanimalsoundswillbebothlowandstableacrossplatforms,andb)animalreactiontothesmallerplatforms

islikelytobelowandconsistent.Thevaryingnoiseoutputfromvesselscanbothmasksoundsandcausechanges

inanimalbehaviourmakingitdifficulttocomparedatacollectedfromdifferentsurveyvessels.Theuseoflarge

vesselsformarinemammalsurveysalsoposestheriskofcollisionwiththeanimalsunderstudy.PoweredAUV

arelesslikelytoplayaroleinpopulationmonitoringduetotheirlimiteddeploymentlifetimesandlikelynoise

interference,particularly forPAMsensors.Althoughrelativelyquietcomparedto researchvessels,propeller

drivenAUVmaydisturbmarinefaunainsensitiveregions,especiallywhenrunninggeophysicalsensorswitha

high acoustic output (Wynn, 2013). The noise generated by the propeller in propeller-driven AUV has the

potentialtomasklowlevelsignalsinthesamefrequencyrange,therebylimitingtheuseofthistypeofAUVfor

PAMpurposes.

Another limitation of the AUV / ASV platforms is their sensitivity to environmental conditions, particularly

currents, and how this affects survey design. Though information was not available for all platforms, the

poweredinstrumentsarelikelytohavemostresistancetocurrents(forexample,theREMUSpropellerdriven

AUVcanholdstationinstrongcurrents).Clearlythedriftingbuoysarethemostaffectedbycurrents,though

theirmajor benefit is their low price compared to other platforms. As discussed in section 11.6.2, current

77


DATE:July2016


researchisfocussingonhowmuchdeviationfromaplannedtrackisneededbeforethesurveydesigncannot

beusedtomakeinferencesaboutanimaldensity/abundanceinthewiderstudyarea.Inthecaseofself-powered

platforms, itmaybe that theextendedsurveyduration is a suitable trade-off for some track linedeviation,

particularlyifthealteredcoursechangesareminorcomparedtotheoverallsurveydesign.

FortheuseofPAMsensorsonAUV/ASV,hydrophonearrayswhichcanlocaliseanimalsmayberequiredfor

detectionprobability.Auxiliarybehaviouralstudiesmaybeneededtoestimatevocalproductionratesand/or

groupsizes.Forsomeregionsdataongroupsizeorvocalizationratesmaybeavailablefrompaststudies,or

might be extrapolated from similar species and regions in the absence of directly relevant data to supply

vocalizationrateandgroupsizevalues.RegardingrangeestimationusingPAM,ship-basedsurveysoftenuse

hydrophonearraysandtheappropriatesoftwaretoestimatebearingstoanimals.Astheshipmoves,multiple

bearingstothesameanimalorgroupofanimalswillcrossovertimetoprovideanestimateofrange(i.e.target

motion analysis). In the case of slowmoving autonomous vehicles, target motion analysis will not work if

animals’positionschangequicklyrelativetothevehicleasbearingstothesameanimal/groupareunlikelyto

cross.Therefore,morecomplexhydrophonearrayswillbeneededthatcanestimaterange,which,inturn,can

beusedtoestimatedetectionprobabilities.However,theversatilityofautonomousvehiclesisanadvantage

here; if amulti-element passive acoustic array (capable of estimating range to a calling animal) cannot be

deployedonasingleplatform,thenmultiplevehiclescouldbedeployedasanarray.Furthermore,amobilearray

couldberelocatedandalsoitsconfigurationcouldbeadjusteddependingonthetargetspecies.Theabilityof

anarraytobereconfiguredwouldbeanadvantageovercurrentfixedarrays,whichoftenhavetohavespacing

forrangeestimationthatisspecies(orgroupsofspecies)specificduetothewiderangeoffrequenciesatwhich

marineanimalsvocalise.

Although an advantage of self-powered platforms is that they can be deployed for long durations, the

consequenceisthattheyareslow-moving.Asdiscussed,thishassomeanalyticalimplications(i.e.theneedto

implementcue-countingorsnapshotmethodstomakesurethatanimalmovementdoesnotcreatebias).Italso

haslogisticalimplications–asmallerareaislikelytobecoveredbyanAUV/ASVthanbyashipinthesame

timeframe.Furthermore,anautonomousvehiclemayneedanadditionalsupportvessel,sothiscostneedsto

befactoredintothesurveybudget.However,despitetheseadditionalconsiderations,thefactthatautonomous

vehiclesbedeployedforextendedperiodsisamajoradvantageformarineanimalsurveys.Byhavingtheability

tothoroughlymonitoraspecificareathroughtime,butalsobeingabletoefficientlymovetootherpartsofthe

studyareamakethemapowerfulassettouse.

8.3.5 Technologiesthatmightbeusedinfuturefieldtrials

Thissectiondescribes(sectionsbelow)andsummarises(Table5)thetechnologiesthatmightbeusedinfuture

trials.

78


DATE:July2016


8.3.5.1 UAS

Allsystemsreviewedinthisevaluationappeartohavethecapabilitiestomonitormarineanimalsinoffshore

regions.Futuretrialsshouldattempttofocusontheversatilityofeachsystem(platformandsensor)indetecting

speciesofvarioussizes,undervariableconditions.SinceUASmayhavedifferentqualitiesandapplications,itis

recommendedtoassesstheconditionstowhicheachsystemmaybemostsuited.Forinstance,forlong-range

experiments,itisnecessarytooperatevehicleswithhigherenduranceandrange,whichmaydemandspecial

flight permits and operating costs. Abundance and distribution studies focusing on the use ofUAS in areas

relevanttoOilandGasrelatedoperations,willprovidenotonlyinformationaboutanimaldistribution,butalso

seasonaldata that canbeused for theO&G industry toconductenvironmentally-consciousdecisions in the

operations conducted (e.g. restricting E&P activities temporally/spatially to avoid sensitive times/areas).

Additionally,studiesusingUASforbehaviouralexperiments,suchascontrolled-exposureexperimentsduring

seismicactivity,canresultinscientificallyvaluableknowledgethatcanalsobeusefulfortheindustry.

Side-by-side comparisons between different autonomous vehicles (AUV, ASV and UAS) can be valuable to

evaluatethedegreetowhichthedataacquiredbythesetechnologiescanbecomplementary.Simultaneous

deploymentsofvarioussystemsmayhelptoestimatedetectionprobabilitiesofagivensystembyusingthe

otherplatformstoconduct“trial”-baseddetectionprobabilityestimation.Thiswould,however,demandahigh

degreeofcontrolovertheconditionsinwhichtheequipmentisdeployed,butitwouldbeneverthelesshighly

relevantforagreaterunderstandingofthecapabilitiesofallthesesystems.

8.3.5.2 AUV/ASV

SeveralPAMandAAMsystemshavealreadybeencombinedwithseveraldifferentautonomousvehiclesand

shown themselves capableof collectingdata.Which technologiesmightbeused ina trialwouldverymuch

dependonthespecificaimsofthattrial.Unlesslogisticalandsafetyconstraintscanbeovercome,itseemshighly

unlikelythatAUVorASVwillbedeployedinthevicinityofaseismicsurveyvesselforreal-timemitigation.For

populationmonitoring, the choice of both vehicle and sensorwould be governedby the specific species of

interest, the geographic area, the water depth, local current strengths, shipping density, are the obvious

questionswhichwouldneedtobeansweredbeforeselectingaplatform,buttherearealsootherfactorssuch

as theexpectedanimalencounter rateand theprecision requiredof the studywhichmightaffectplatform

choice. Given PAM systems’ inability to distinguish some species groups (e.g. accurately separating dolphin

whistlestospecies),itwouldbenecessarytodeterminewhatlevelofspeciessubdivisionwouldberequired.

Answerstothesequestionswouldguidethetypeofsensorrequired,thedurationforwhichmonitoringmaybe

requiredandthenumberofvehicles(orlisteningstations)thatshouldbedeployed.

AUVandASVmayalsobesuitabletools forsoundmeasurement,particularly inremoteareaswheredata is

hardertoobtain.UnderwaterbuoyancyglidersandtheWaveglideraresuitableforthistaskifavesselcandeploy

79


DATE:July2016


themnearby.Alternatively, theAutonaut inparticularwouldbewell suited to this taskdue to longmission

durationandindependence.

Table5.Recommendedtechnologytypesthatmightbeusedinfuturefieldtrialsforthedifferentmonitoringtypes.Themonitoringtypesarefurthermoredividedintomitigationmonitoringinareaseitherclearfromorbusywithotheroperationalgearortraffic,short-term(hours,days)orlong-term(weeks,months)monitoringandfocal-followsconductedwithstaticormobilesystems.L-t-a=Lighter-than-airaircraft.

Monitoringtype

Mitigation Population Focal-follow

Sensor Vehicle clear busy short-term long-term static mobile

Electro-optical Poweredaircraft ALL ALL

Bramorc4Eye,

BramorgEO,

BramorrTKFulmar,

Jump20,PenguinB

andScanEAgle

NONEALL–Dependenton

thesensortechnology

Motorisedgliders NONE NONE NONE NONE NONE NONE

Kites ALL NONE ALL NONE ALL ALL

L-t-aaircraft ALL NONE ALL ALL ALL ALL

PAM

Powered

AUV NONE NONE NONE NONE

Dependsonspeedof

animalsandvolume

forwhichtrackingis

required.

ASV

C-Worker,C-

Enduro.

Scoping

aheadof

source,

specificallyfor

LFdetections.

NONEC-Worker,C-

Enduro.NONE

Self-

powered

AUV NONE NONESlocumand

Seaglider

Slocumand

Seaglider

ASV NONE NONE Waveglider Waveglider

Drifter NONE NONE DASBR NONE

AAM

Powered

AUV NONE NONEALL–Dependentonthespecificspeciesofinterest,the

geographicarea,thewaterdepth,localcurrentstrengths,

shippingdensity,theexpectedanimalencounterrateand

theprecisionrequiredofthestudyandpossiblymore.

e.g.Bluefin9/9M,A9MandREMUS100forshallowand

mediumwatersandshortdurationsurveys;Bluefin12Dand

REMUS600fordeepwaterandlongdurationsurveys;

Bluefin21,A18DandREMUS6000forverydeepwatersand

longdurationsurveys.

ASV

ALL*

*see

population

monitoring

NONE

Self-

powered

AUV NONE NONE

ASV NONE NONE

Drifter NONE NONE All NONE

8.3.6 Datagapsandrecommendations

This section provides a description (sections below) and summary (Table 6) of the data gaps and

recommendationsidentifiedinthisreport.Oneoverarchingrecommendationdirectlyconnectedtothisreview

istheestablishmentofacomputerbasedexpertsystemthatusesthevastamountofinformationgatheredin

80


DATE:July2016


this review and the preceding review on low visibilitymonitoring techniques as published in Verfuss et al.,

(2016).Thisexpertsystemshouldbebasedontheevaluationcriteriaestablishedinbothreviews,andshould

providealistofthemostappropriatesystemsandmethodsasanoutputbasedonthemonitoringrequirements

fedintotheexpertsystem.Acost/benefit-analysistoolcouldalsobeimplemented.Thissystemwouldneedto

bebasedonalivedatabasethatwillbeupdatedonaregularbasistokeeptheinformationfromthisrapidly

developingfielduptodate.

8.3.6.1 UAS

Asnewsystemsarise,additionaltestingisrequiredtofurthervalidatethequalityofthedataacquiredduring

unmannedaerialsurveys.Side-by-sidecomparisonswithcurrentmonitoringmethodsstillrepresentalargedata

gap, which are of high relevance particularly when considering the replacement of current methods.

Additionally,theamountofinformationprovidedbyUASproducersislimited,whichrestrictstheinformation

providedinthisreport.However,evenwithlimitedinformation,itisalreadypossibletoseethatthesystems

haveahighpotentialforapplicationtooffshoremonitoringofmarinemammals,turtles,andfish.

Thoughthesystemsshowgreatpromise,thedevelopmentofautomaticdetectionalgorithmsapplicabletothe

variety of data gathered using UAS, is one of the key areas that requires further development. Detection

algorithmsneedtobeadaptabletothevaryingenvironmentalconditionstowhichtheyareapplied,andthe

numberoffalsedetectionsneedstobewellquantified(andminimisedforreal-timemitigationpurposes).Real-

timedatatransfermayimprovedatafilteringandacquisition,butitremainsdependentonahumanobserver

toverifyananimaldetection,whichmaynotalwaysbeeasyiftheaircraftisflyingathighspeed.Finally,there

isatrade-offbetweenthedegreeoffalsepositivesandtheprobabilityofmissingadetection–amoresensitive

algorithmwillallowahigherdegreeoffalsepositivesbutwithalesssensitivealgorithmthereisahigherchance

ofmissingadetection.Foranyparticularapplication,thebalancebetweenthesemaydifferdependingonthe

consequenceofmisseddetections–forexampleinreal-timemitigationthereislikelytobemoreofaneedto

implementamoresensitivealgorithmthanforpopulationmonitoring.

Asplatformsandsensorsvaryintheircapabilities,itwouldbeworthassessingthepotentialuseofeachtypefor

specificapplications.Forinstance,testingthedifferenceindataqualitybetweenvideoandIR,orstill-photoand

IR,particularlywhenmonitoringlargeareasandpresenceofsmallspecies.Thedetectionofmarineturtlesand

largefishshouldalsobeunderfocus,sincemostmonitoringstudiesusingUASinvolveeitherterrestrialmapping

ormarinemammaldetections.

ThoughtherehasbeenanincreaseintheuseofUASinmarinemammalresearch,relativelyfewstudieshave

reported their effects onmarine animal behaviour (Smith, 2016). Altitude plays a key role in assessing the

potentialeffectsofUASonmarineanimals,butithasnotyetbeenestablishedwhetherthereisadistinction

betweendisturbancesfromnoiseversusvisualdisturbanceasafunctionofaltitude(Smith,2016).Asaltitude

decreases,itismostlikelythatthetriggersforbehaviouralresponseschangefromacoustictobothacousticand

81


DATE:July2016


visual ones. For cetaceans in general, a behavioural disturbancemay be triggered during low altitude UAS

operations.Whencomparedtomannedaircraft, severalstudieshavedocumentedthatUAScauseasmaller

degreeofdisturbanceandavoidancebehaviour(Acevedo-Whitehouse,2010;Mulaca,2011;Moreland,2015).

ThisispossiblyasthenoiselevelsproducedbyUASarefarbelowthosefrommannedaircraft(Jones,2006;van

PolanenPetel,2006;NMFS,2008;Hodgson,2013;Goebel,2015;Pomeroy,2015).Overall,asUAStechnology

improvesweare likelytoexpectafurtherspike inUASoperationswithanapplication inanimalmonitoring,

particularly inmarineregions. It is therefore relevant tounderstandtheeffectsof this technology itselfand

comparedwithcurrentmonitoringmethodsforvariousspeciesandlocationsbeforeitisacceptedtoreplaceor

complementothermethodsusingaprecautionaryprinciple.

The evaluation presented here provides a glimpse at how UAS may efficiently collect valuable data and

transformoperationsfortheoilandgasindustry.ThevalueofUASforreducedriskandincreasedcostavoidance

isvisibleacrossthetestsconductedovertimeandtheprogressofthistechnology.LowerUASoperatingcosts,

for example, allow more frequent aerial inspections resulting in enhanced awareness and more proactive

maintenance.AsindustriesattempttofullyunderstandthebenefitsofemployingUAS,theyrequirethorough

planningtoutilisethebesttechnologytomeettheirneeds(i.e.takeinconsiderationregulatoryandtechnology

updatesthatmayaffectdataacquisitionandanalysis),andeffectivelyharnesstheincreasedvolumeofdatato

makeinformeddecisions(Gibbens,2014).

8.3.6.2 AUV/ASV

Thereareanumberofplatformsandsensorsatahighstateoftechnologicalreadinesswhichcanbeusedfor

thecollectionofPAMandAAMdatafromautonomousplatforms.Withalittleeffort,manyofthesesensors

couldbe combinedwithmanyof theplatforms and they couldbe sent out to collectmonitoringdata. The

problemsthatlimittheuseofthesesystemsalsooftenapplytosurveysconductedfromshipsinthatitisdata

interpretationratherthandatacollectionwhichisthefundamentallimitation.Theseproblemscanhoweverbe

moreacuteonsmallerlowerpoweredautonomousvehiclesthantheyareonshipsforanumberofreasons.

For PAM, the current trend is to deploymore sophisticatedhydrophone arrays that can at least determine

bearingstosoundsources.Fastmovingvesselscanalsousetargetmotionanalysistodeterminerangetosome

species, and the possibilities of combining visual and acoustic methods on the same platform can help to

understandoveralldetectionprobability.ThedevelopmentofPAMsystemsforautonomousplatformsthatcan

employmoresophisticatedarraysthereforerequiresconsideration.

AlthoughthedevelopmentofdetectionalgorithmsisfurtherdevelopedforPAMthanforvideoandAAM,in

mostsituations,thereliabilityofthesealgorithmsmeansthathumanobserverscontinuetoplayanimportant

roleindataanalysiswhetheritbeformitigationorpopulationmonitoring.Whiledetectionandclassification

algorithmswill always continue to improve, itmust be understood that theywill never be perfect and are

unlikelytoeverprovidesimplered/greenstop/gotrafficlightlikesystemsformitigationinmostsituations.This

82


DATE:July2016


fundamentallimitationshouldbetakenintoaccountaswemoveforwardwithdevelopingthesesystemsand

whilewewouldnotsuggestthatfurtherresearchintoalgorithmsforautomaticdetectionisnotworthwhile,

research into both themagnitude and the effects ofmis-detection andmis-classification and investment in

systemswhichmakeitpossibleforhumanstocontinuetoplayaroleindecisionprocessesisprobablyofhigh

importance.

AAMfuture researchshould focusondevelopingdetectionalgorithmsanddatacompression techniques for

returningsummarisedand/orrawdataoverlowbandwidthcommunications.Simultaneousapplicationofthe

differentsensortypesmayhelpquantifyfalsedetectionratesandestimatedetectionprobabilitiesfordifferent

platform/sensorcombinationasdiscussedinsection8.3.5.2andasoutlinedinVerfussetal.(2016).Asforall

platforms,well-annotateddatasetsthatcanbeusedtotestdetectionandclassificationalgorithmsareessential.

TheperformanceofthedetectorsusedwithAUVandASVwillalsoneedtobetested/proveninthepresenceof

localindustrialorambientnoisesources.

TheuseofPAMsensorsformarineanimalmonitoringinrecentyearshashighlightedtheneedforcontinued

behaviouralresearchtobetterunderstandthecontextsinwhichanimalsproducesound(Marques,2013).This

iskeytoinferringanimaldensityfromacousticdata.Optimalsurveydesignforpassiveacousticsurveysisalso

anongoingareaofresearch(Marques,2013)but,asmentionedinSection8.3.5.2,theseinvestigationswillrely

onspecificsurveyscenariosinwhichtotestdifferentconfigurationsofplatformsandsensors.

Ingeneral,behaviouralinformationonmanyanimalspeciesformsahugedatagap.Therefore,auxiliarystudies

onanimalbehavioursuchasthevocalproductionrates,groupsizes,divingbehaviourorsurfacingrateswillhelp

toquantifydetectionprobabilitiesofthetargetspeciesandreducethesurroundinguncertainties.

8.3.6.3 Operationalaspects

Health and safety is the highest priority with regards to the operation of autonomous vehicles. Technical

reliabilityforthesesystemsisakeyissueforanycraftthatisconsideredformarineanimalmonitoring,andnot

allautonomousvehicleshaveaproventrackrecordoftheirtechnicalreliability.Anyconsideredcraftshould

thereforebetrialledextensivelytoensuretheirreliabilityduringoperationtoensuresufficientsafety.

FurtherpromotionoftheHSEprocedureswouldbethedevelopmentofanindustry-wide“codeofpractice”for

operation of autonomous vehicles in industrial operational fields to ensure safe interaction with other

marine/airspaceusers.Therearemanyissuesconcerningtheregulations,whichdifferamongcountriesandlack

internationalharmonisationonbothstandardsandregulations.Thiscancreatesomeresistancefromcurrent

airorwaterspaceusersandpublicapprehension.Thisshould,therefore,befurtherdevelopedinordertocreate

asmoothertransitionto,andacceptanceof,autonomousvehicleuse.Therefore,onerecommendation is to

constructaframeworkgoverningrulesandregulationsthatareflexibleandamendabletothespecificsofan

industrialprojectandtotherapiddevelopmentofautonomousvehicles.Thisframeworkshouldallowforthe

83


DATE:July2016


rangeofapplicationabilitiesofautonomousvehicles.Afurtherstepwouldbetheinternationalharmonisation

onbothstandardsandregulations.

Further research into “detect-and-avoid” systems for autonomous vehicles is another step towards safe

operation of autonomous vehicles. There are, for example, safety and regulatory issues associated with

operatingaUASincivilairspace,whichneedtobeaddressedbytheaviationandairspaceauthorities inthe

countrieswhereunmannedsurveysaretobeconducted.Thelackofsystemswithsee-and-avoidtechnology

thatwouldpreventtheUASfromcollidingwithanaircraftlimitstheabilitytoconductstudiesinremoteareas

andisalsooneofthemainconcernsraisedbyaviationauthorities.Themandatoryinclusionofsee-and-avoid

technologyinUASisunderdiscussionatthemomenttohelpachievetechnologyacceptanceandsafety.

84


DATE:July2016


Table6.Shortlistofrecommendationswithmedium(M)orhigh(H)priorityandwithmedium(M)orhigh(H)urgency,alongwiththerelatedtopicanddatagap.

Topic Recommendation Priority Urgency Datagap/issue

ALL

Studies comparing MMO/PSO, PAM, UAS, AUV, ASV and tagging to assess the

capabilitiesofeachsystems,determinetowhichextenttheymaycomplementeach

other,andprovideaccuracyinanimaldistributionandbehaviourstudies

H M

Need for assessment of technology

capabilities in relation to all monitoring

methods

AAMFocus on developing detection algorithms and data compression techniques for

returningsummarisedand/orrawdataoverlowbandwidthcommunicationsM M

Detection algorithms and data compression

techniquesnotwelldeveloped

PAMDevelopment of PAM system for employing sophisticated arrays for bearing and

distancemeasurementM M Limitationindatainterpretation

PAMTestingofperformanceofsystemsinthepresenceoflocalindustrial/ambientnoise

sourcesM M ApplicabilityofPAMinanoisyenvironment

Behavioural

data

Improve knowledge of target species behaviour through dedicated behavioural

studiesH H

Multiplierssuchasgroupsize/callproduction

rates are often important in population

monitoring and also relevant for mitigation

monitoring.

Operational

safetyExtensivetestingofsystemstoensurereliabilityduringoperation H H Reliabilityofsystemsoftennotknown

Operational

safety

Developmentofanindustry-wide"codeofpractice"toensuresafeinteractionwith

othermarine/airspaceusersM M

No"codeofpractice"existing for interaction

between autonomous vehicles and other

marine/airspaceusers

Operational

safety

Constructionofaframeworkgoverningrulesandregulationsthatareflexibleand

amendable to specifics of industrial project and development of autonomous

vehicles

M M

Regulations differ among countries, lack

international harmonization on both

standardsandregulations.

Operational

safety

Research into collision risk and integration of "detect-and-avoid" systems into

autonomousvehiclesforsafeoperationofautonomousvehiclesH H

Integration of "detect-and-avoid" systems

oftennotgiven.

Sensors Improvementofalgorithmsfor(real-time)detection,classification,localisation M MDetectionalgorithmformarineanimalspecies

notalwaysimplemented,lackofclassification

85


DATE:July2016


Topic Recommendation Priority Urgency Datagap/issueof many species, localisation option may be

missing

Sensors

Identification of the magnitude and effects of miss-detections and miss-

classifications.

Simultaneousapplicationofdifferentsensortypestoquantifyfalsedetectionrates

andestimatedetectionprobabilitiesfordifferentplatform/sensorcombinations

H HFalsealarmratesandrateofmissesnotwell

known.

UAS

Furtherstudiesfortechnologytestinginoffshoreregionsbefore,duringandafter

seismicoperation.Comparisonwithcurrent(manned)methodsinassessinganimal

distributionanddensityinareasofinterestforthepetroleumindustry

H M

Technology capabilities in relation to current

aerial and ship-borne mitigation monitoring

methodsnotknown

UAS

UsecurrentavailableUASdatatodevelopversatiledetectionalgorithmsthatcanbe

used both during real-time transmission and for post-processing of the data

acquired.

H HAutomaticdetectionalgorithmsgenerallynon-

existingoratbasiclevel.

UASDevelopmentofreal-timedetectionsystemsandsupportforfurtherdevelopment

ofexistingreal-timedetectionsystems.H H

Hardly any real-time detection systems

existing.

UASInvestigatefurthercapabilitiesofUASthatcanbecombinedwithsurveillance,such

asdeploymentofotherequipmentatseaM M

Need for further investigation of the

capabilitiesofsensorandaircraftsystems

UAS

Compare different sensor systems for species such as fish and marine turtles,

tackling both the testing of sensor types and need for more knowledge about

sensitivespeciesinregionsofinterestforthepetroleumindustry.

M MNeed for further investigation of the

capabilitiesofsensorsystems

86


DATE:July2016


9 Further information 9.1 Requirementsofautonomousvehiclesformarineanimalmonitoring

Theoceancanbeahostileenvironmentandtheuseofautonomoussystemswillnotonlydecreasetheriskto

surveypersonnel,butalsoenablesurveyingofhard-to-reachstudyareas,suchasareasaffectedbyicecover

(Dickeyetal.,2008).Theuseofautonomousvehiclesmaybemorecost-effectiveandabletogeneratemore

robustsurveydataandresultsthantraditionalshipandaerial-basedsurveys.Dependingonthesensorsused

withanautonomousplatform,round-the-clockmonitoringmaybepossible,incontrasttomannedsurveysthat

collectvisualsightings,whichcanonlybeconductedindaylighthours.Theuseofautonomousvehiclescreates

opportunitiestostudytheverticalvariabilityintheworld’soceans(usingdivingAUVsystems),whiletraditional

visual survey methods can be ineffective at detecting small aggregations of animals that spend significant

periodsoftimesubmergedandoutofviewattheseasurface(Baumgartneretal.,2013).Ingeneral,autonomous

systemshavegreatpotentialtoimprovethespatialandtemporalcoverageofmarineanimalsurveys(e.g.Van

Parijsetal.,2009).

Thereareseveralexistingexamplesoftheuseofautonomousvehiclesformarineanimalsurveys,whichare

summarisedinsection9.2aswellasTable7andTable8.However,itisimportanttoassessthesetechnologies

andconsiderwhethertheinstrumentsarecompatiblewithexistinganimalsurveymethodologies,orwhether

newsurveyingmethodsarerequired.

Thereisawiderangeofautonomousvehicles(outlinedinsections7.2),thus,theircapabilitiestocollectsuitable

data formarineanimalmonitoringwill vary. This sectionwill provideageneraldiscussionofmarineanimal

monitoring, identifying key features of survey design, data collection and data analysis which need to be

considered inrelationtousingautonomoussystemsasmonitoringplatforms.Thesefeatureswillbeusedto

assessthepotentialsuitabilityofthevariousautonomousvehiclesformonitoringmarineanimalsfortheO&G

industry.

Thekeyconsiderationswhenassessingwhetherautonomoustechnologiesaresuitableformitigationmonitoring

and surveying marine animals for both population-level and individual-level studies, can be split into the

followingtopics:

• Surveydesign

• Datacollection

• Dataanalysis.

In this section, each topic will be discussed in relation to autonomous technologies to understand which

capabilitiesanautonomoussystemneedstocovertoberegardedassuitableforacertainmonitoringtask(see

87


DATE:July2016


alsosection8.3.1).Examplesonthecurrentuseofautonomoussystemsusedformarineanimalmonitoringcan

befoundinsection9.2.


9.1.1.1 Surveydesign

Intermsofmitigationmonitoring(section7.1.1),thesurveyneedstobedesignedsothatthemonitoringensures

a timely detection of a target animal within the monitoring zone. Continuous real-time or near real-time

monitoring is required to achieve a timely detection so that the time between a target animal entering a

monitoring zone, its detection and the subsequent mitigation decision is sufficient to enable an effective

mitigationcascade(seeVerfussetal.,2016forfurtherdiscussion).

It is also important that the probability of detecting a target species is high. Key to maximising detection

probabilityresults inthechoiceoftheappropriatesensortechnology.Electro-opticalsensorscombinedwith

aerialvehiclescanonlydetectanimalspresentabove,at,orclosetotheseasurface(anexceptionisincaseof

thermal IR,which cannot detect sub-surface animals), PAM sensors combinedwithAUV / ASV require that

animalsvocaliseandthatthosevocalisationsaredistinguishablefrombackgroundnoiseinordertobeableto

detectthem,andAAMsensorsonlyworkunderwater.Anextensiveevaluationofwhichtargetspeciesisbest

detectedwithwhichsensortype,andthe influenceofenvironmental factorsonthedetectionprobability, is

giveninVerfussetal.,(2016)andwillthereforebeomittedinthisreview.

Asecondconsiderationistheresolutionofthespatialandtemporalsurveycoverageofthemonitoringzone.

Thecoverageneedstobesuchthatthechancesofdetectingananimalpresentwithinthemonitoringzoneare

maximised.

Theabilityofasystemtoclassifyand/orlocaliseananimalisalsouseful.Classificationand/orlocalisationisnot

necessarily essential for mitigation monitoring but will decrease the likelihood of mitigation processes

conductedbasedonfalsealarms.Theabilitytoclassifydetectionssothattargetspecies(orspeciesgroups)can

bediscriminatedfromnon-targetspecieswilldecreasetheamountoffalse-detections.Regardingthelevelof

classificationanddetailrequiredaboutagivendetection,inmanysituations,speciesidentificationiseithernot

requiredatall,or isonlyrequiredtothe levelofbroadspeciesgroups(e.g.adolphinora largewhale).The

sensortypesincludedinthisreviewhaveclassificationabilitiestosomedegree.Anextensiveevaluationofthis

matterisgiveninVerfussetal.,(2016),andwillbereconsideredinthediscussionofthisreview(section8.3.1).

Theabilitytolocaliseatargetanimalwillalsodecreasetheamountoffalse-detections,giventhatmitigation

measures will only be taken when a target animal is present in themonitoring zone. Animals outside the

monitoringzoneshouldeithernotbedetected,oritshouldbepossibletolocalisethemwithsufficientaccuracy

tobesurethattheyareoutsidethezone.Localisationdependsontheconfigurationofasensorsystem.Electro-

opticalsystemswillneedtohave(ataminimum)referencepointstoallowthemonitoringzonebordertobe

88


DATE:July2016


identifiedontherecordedimages.Inmostcircumstances,singlehydrophonePAMsystemscannotdetermine

range.Smallarrays,asmightbedeployedfromasinglevehicle,candeterminebearingtosoundsourcesandfor

some species,multiplebearings fromdifferentpoints along the trackline canbeused to estimate range. In

certainwellunderstoodpropagationconditions,rangecanbedeterminedusingmultipatharrivalsofsignals(e.g.

surfaceandbottomechoes)ordispersivemodepropagationmodels.MostAAMsystemshaveexcellentrange

determination capabilities and high angular resolution. Further discussions on this matter can be found in

Verfussetal.,(2016).

9.1.1.2 Datacollection

Forrealtime,ornearreal-timeapplications,suchasmitigationmonitoring,considerableamountsofdataneed

tobetransmittedtoanoperatorifaccuratedecisionsaretobemade.Inalimitednumberofcases,thishas

been possible using low bandwidth satellite links where summary information or short sound clips from

automaticdetectorsaresenttoshoretobecheckedbyahumanoperator.Formostoperationalscenarios,the

quantities of data that need to be presented to an operator are currently too large to send through low

bandwidth.However,forshortrangecommunication,freespectrumwirelesssystemscanbeusedtosendhigh

bandwidthdataoverrangesofseveralkilometres.Acousticprocessing isanactiveareaof researchand it is

possiblethatforsomesituationstheamountofdatathatneedtobesentinorderthataneffectivemitigation

decisioncanbemadecouldbesignificantlyreduced.

Formitigationmonitoring,themostpracticaloptionisthereforetouseafreespectrumradiomodemlinkto

senddatatoanearbybasestation(whichmightbeononeoftheoperationalvessels)whereanoperatorcan

viewandverifydetectiondata.

9.1.1.3 Dataanalysis

Formitigationmonitoringitisadvisabletohaveatrainedhumanoperatortocheckdetectionsidentifiedbya

system’s detector before mitigation procedures are implemented. Detectors can be used to reduce data

volumesthatneedtobestoredortransmitted,butnonehaveyetbeensufficientlyvalidatedasbeingcapable

ofaccuratelymakingacompletedecisionconcerninganimalpresencewithahighdegreeofcertainlyandlow

falsealarmrate.Detectorperformancemayalsobecomepoorinthepresenceofnoise(industrial,biological,

weathergenerated,etc.).AutomaticdetectionandclassificationisgenerallymoreadvancedforPAMthanfor

AAMandopticalmethods.However,therearenosystemsofanytypewheredetection issoautomaticthat

automaticdetectorscanberelieduponontheirown.

89


DATE:July2016


9.1.2 Animalpopulationmonitoring

9.1.2.1 Surveydesign

Asoutlinedinsections7.1and11.6,surveydesignisanimportantfeature,firstly,tobeabletoinferdensity

and/orabundanceinthestudyareaand,secondly,fordetectionprobabilityestimation.Therefore,autonomous

vehiclesthatcanadheretodesignedsurveytransectlines,orremainstationarytocreateamonitoringpoint,

wouldbeparticularly suited for acquiring surveydata. Technical termsused in this sectionareexplained in

section11.6.

Ifpilotstudyinformationisavailableontheexpectedencounterrateintheregion,theamountoflinelengthor

thenumberofpointsneededtoachieveanacceptablelevelofvarianceintheresultscanbeestimated(Buckland

etal.,2001).Formovingsystems,thepowerandspeedofthesystemwillneedtobeconsideredtoensurethat

it is logistically possible to complete the survey in the required timeframe. Deployment duration and

communicationrangewiththegroundstationshouldbeofsufficientlengthtoallowforsuccessfuldeployments

fromandreturnstolandorasurveyvessel(asnotedbyKoskietal.,2009ainareviewofUAS).

If systemswere tobeusedas stationarymonitoringpoints then, ideally,enoughsystemswouldneed tobe

available toachievesimultaneousmonitoringatallpoints. Itwouldbepossible to split thesurveyarea into

smallerblocks tosurveywith fewersystemsatanyone time,but thentheability to teaseapart spatialand

temporalfactorsthatmayaffectanimaldensity/abundancewouldbereducedorlost.

Duringthesurveydesignstage,everyeffortshouldbemadetokeepthenumberofmultiplyingparameters(or

“multipliers”-seesection11.6forexplanation)requiredtoestimateabsoluteanimaldensityorabundancetoa

minimum.Foreveryextraestimatedmultiplier,suchasdetectionprobability,additionaluncertaintyisaddedto

theanalysis, increasing theoverall varianceestimates. Furthermore, there is thepossibility thatamultiplier

estimate may be biased (e.g. an incorrectly applied call production rate), which may result in biased

density/abundanceestimates.Therefore,anidealdesignwouldbeonewhereallanimalsinthemonitoredarea

were certain to be detected and individual animals could be counted (so no group size or cue production

multiplierwouldberequired,forexample).Detectionprobabilityestimationmaynotalwaysberequired,ifit

canbeassumedthatallavailableanimalswithinasurveyedtransectorpointaredetected(asmaybethecase

inUASsurveyswhererecorded imagesarecollectedandanalysedpost-survey).However,mostpopulation–

level surveyswill require some formofmultiplier so substantial effort should bededicated to planning the

optimalsurveygiventhestudyarea,thestudyspeciesandlogisticalconstraints.Itisimportanttonotethatthe

surveydesignrequiredtoachievecertain,orashighaspossible,detectionprobabilitiesforreal-timedetection

withtheaimofmitigationmonitoringwillbedifferentfromthesurveydesignforpopulationsurveys(Verfuss,

etal.,2016).

Foreveryrequiredmultiplier,anappropriateestimationapproachshouldbeplannedatthesurveydesignstage,

andmayinvolvethecollectionofauxiliarydata.Probabilityofdetectionshouldbeestimatedusingoneofthe

90


DATE:July2016


“standard”densityestimationmethods,suchasdistancesamplingorSECR,sothedesignofthesurveyshould

ensurethatappropriatedataaregeneratedfortheseanalyses(i.e.,estimatedhorizontalrangestoanimalsfor

lineorpointtransectsampling,orassociatingthesameacousticeventsacrosshydrophonesorre-identifying

individuals across surveying occasions for SECR). Particularly forUAS technology, implementingmethods to

accountformissinganimalsdirectlyonthetransectlineoratthepointduetonon-constantavailabilityshould

be considered (seeChapter11,Bucklandet al., 2015 fora recent reviewofmethods toaddress this issue).

AltitudeandspeedoftheUASwillbeofimportancewhendeterminingtheavailabilitybias;thetemporaland

spatialfieldofviewoftheUASmaybetoonarrowtoallowaseconddetectionofananimalwhenresurfacing,

soalternativeapproachesmayhavetobeconsideredfordeeperdivingspecies.InthecaseofAUVtechnologies

thatcollectpassiveacousticdata,itislikelythatcallproductionrate,orproportionoftimeanimalsarevocally

active,willberequiredasamultiplier.Theexactmetricwilldependontheexactformofthedensityestimator.

Therefore,anassessmentoftheavailabilityofsuitabledatafromtheexistingliteratureneedstobeconducted,

orassociatedbehaviouralstudieswillneedtobeplannedaswellas themainsurvey.Giventhatbehaviour-

linkedmultipliersarelikelytobecontext-specific,itisrecommendedthatmultipliersareestimatedatthesame

timeandplaceasthemainsurvey(Marquesetal.,2013a).

Furthermore, if it is suspected thatmodel-based inference (see Section11.6.2)will be required toestimate

abundanceinthewholestudyarea(e.g.ifthechosenvehiclemaydeviatefromitsplannedtrack),ordensity

surfacemodellingisrequired,thenappropriatecovariatedatasetswillhavetobeobtainedfromotherexisting

datasetsorcollectedfromthestudyarea.

GiventhenumberofavailableAUVandUASplatforms,itispossiblethattheremaybeseveraloptionsatthe

surveydesign stage.Different combinationsof systemand sensormaybe suitable for the samesurvey. For

example, smaller animalsor animalswith little surface contrast caneitherbedetectedwithhigh-resolution

camerasmountedonaUASoperatingatahigheraltitudeandspeed,orwithlowerresolutioncamerasoperated

bysmallerUASflyingataloweraltitudeandspeed.Thesizeoftarget,lightconditions,seastate,operational

altitude, speed and contrast between target and background are also important factors to consider when

determiningwhich imaging system is required for an operation. All survey design optionswill likely involve

differentprosandconsthatwillbesurvey-specific.

In order to assess animal population sizes, real-time detection is not required. However, there would be

advantagestosystemsthatcouldrelaydatabackinreal,oralmost-real,time.Forexample,surveydesignscould

beupdatedoradapted,dependingonwhatspeciesweredetected–thismaybeveryusefulifmultipleAUV/

ASVinstrumentsweredeployedtoformanarray,whichcouldhavealteredspacingdependingonthetypesof

speciesbeingacousticallyencountered.

91


DATE:July2016


9.1.2.2 Surveydatacollection

Regardlessoftheparticularmethodologicalprotocolthatasurveymayfollow,therearebasicdatacollection

requirements foranysurvey.Firstly,a temporalandspatial recordof thesystem’ssurvey route isessential.

Secondly,systemsshouldcollecteithervisualoracousticdatathatcanbeusedtoidentifyspecies.Inthecase

ofUASsurveys, stabilizationof thecamera,especiallywith regard topitchand rollof theaircraft,hasbeen

identifiedasbeingimportant,particularlywhenusingvideoforthedetectionofmarinemammals(Koskietal.,

2009a).AUVandASVsystemsthatuseon-boardprocessingofacousticdataoughttokeepexamplesofrawdata

sothatfalsedetectionproportionscanbeassessed.

Clearly,whateversystemisused,inmostsituationsitisdesirabletostoreandpresenttoacompetentoperator

asmuch rawdataaspossible if accuratemanagementdecisionsare tobemade.For long termmonitoring,

considerableamountsofdatacanbestoredonhighcapacitymemorycardsorharddrivesand it isonlythe

lowest power and long duration vehicles (such as submarine gliders) that require significant real-time data

reductionforthisapplication.Ourrecommendationfor longtermmonitoringistostorealldata,orasmuch

dataasispracticallypossibleandtosendaminimalamountofdatatoshoreprimarilyasasystemstatusand

healthcheck.Thedetaileddataprocessingcanthenbeconductedoncethesystemhasbeenrecovered.

9.1.2.3 Surveydataanalysis

Thedevelopmentofalgorithmstodetectmarineanimalsautomaticallyfromaerialimagesisanareaofactive

research (Ireland et al., 2015). Image processing for fisheries stock assessments is also being investigated

(Mellody,2015).Therefore,atpresent,thevastmajorityofrecordedsightingsdatawillbemanuallyanalysed

byanoperator,eitherinnearreal-timeorafterthevehicleisrecovered.Inanyautonomousvehiclesurvey,the

particularabundanceestimationmethodologyusedwilldeterminewhatanalysesarerequired.Forexample,

estimationofhorizontalperpendiculardistanceforlineorpoint-transectsampling,individual(re)identification

forspatially-explicitcapture-recapture(SECR)andgroupsizeestimation(seesection11.6forfurtherdetails).

Where multiplier estimation analyses are required, such as group size estimation, collected data can be

subsampledifsamplesizesareprohibitivelylarge.However,asystematicrandomsubsample(i.e.,evenlyspaced

sampleswitharandomstartpoint)ofthedatamustbetakensothatthesubsampleisatruerepresentationof

thefulldataset.

Thedataanalysisconsiderationsoutlinedabovearenotsignificantlydifferentfromcurrentconsiderationsfor

collecting visual or acoustic data for wildlife surveying. However, the differences of UAS/AUV platforms to

currentlyusedplatforms(i.e.,ships,aircraft,ormooredacousticinstruments)meanthattherearesomedata

analysisconsiderationsrelatingtoautonomousvehiclesthatarecurrentareasofresearch.Thesearediscussed

here,thoughdefinitiveconclusionsand/orrecommendationsarenotyetavailable.

Thefirstissuerelatestosurveyeffortandanimalmovement.Typically,surveyeffortisdefinedaslinelength

travelled for line transect surveysand time spentmonitoring forpoint transect surveys.However, it canbe

92


DATE:July2016


considered that slowmovingplatforms,although theymovealong transect lines, areessentially completing

point transect surveys,where thepointmoves slowly through space. The slowmovement of platformshas

implicationsfor(1)howsurveyeffortisdefinedinthedensityestimatorand(2)whetheranimalmovementmay

beproblematicornot(Glennie,2015). Ifaplatformmovesconsiderablyslowerthanthestudyanimals(or is

stationary),anybiascausedbyanimalmovementcanbeavoidedbyeithercountingcues(e.g.acousticcues,or

visualcuessuchaswhaleblows)ortreatingthesurveyasaseriesof“snapshot”momentsthroughtime.The

lattermethodworksbyestimatingdensityinshortperiodsoftime(i.e.,asnapshot)whereanimalmovementis

deemed tobeminimal (e.g.Ward,2012;Hildebrand,2015). Surveyeffort is thendefinedas thenumberof

snapshotsusedintheanalysis.Currentresearchisinvestigatingtheeffectofsnapshotwindowlengthondensity

estimation using data from slowmoving autonomous vehicle, specifically gliders and drifting platforms. In

addition,workiscurrentlyongoingtodeterminehowmuchdeviationfromaplannedsurveyisneededbefore

model-basedinferenceisrequiredtoestimateabundanceinthewholestudyarea.

9.1.3 Focalanimalstudies

Inthecaseoffocalanimalstudies,surveydesign,datacollectionandanalysisneedtobespecificallytailoredto

thestudyobjectives.Wethereforehighlightsomegeneralkeyconsiderationsinthissection.

Autonomousvehicleswillneedtobeeither(1)mobileinordertofollowanindividualofinterest(viamanualor

automaticpiloting)or(2)remainstaticwithafieldofviewsuchthatananimalcanbemonitoredforaperiodof

time.Whenusingamanuallypilotedsystem,real-timedetectionfeedbackisnecessaryforthepilottotrackthe

animal.Automaticpilotedsystemswouldneed tohaveadetect-and-track routine thatallows forautomatic

trackingofadetectedanimal.Mobilesystemswouldadditionallyneedtohaveageo-referencingsystemthat

allows thedeterminationof the target animal’s location.Anassociated time-depth recordermayalsobeof

interestforobtainingdive-profilesoftheanimal.

Duringthefocalfollow,detectionprobabilityshouldremaincertain,sothatacompleterecordoftheanimal’s

behaviourduringthefocalstudyisrecorded.This ismoredifficulttoachievewithPAMsystems,becausean

animalisonly‘observable’whenitisvocalising.Theabilitytoperformfocalfollowswillalsobelimitedbythe

communicationsystemandtheoperationalrangeofthesystem.

93


DATE:July2016


9.2 Currentknowledgeonautonomousvehicles

9.2.1 Independentsystems

9.2.1.1 Unmannedaerialsystems

UAS have recently been introduced as alternative platforms to overcome the limitations of manned-aerial

surveys formarinestudies,with theultimatebenefitbeingtheirability toperform"dull,dirty,dangerous10"

tasksmoreefficientlyandatafractionofthecostofmanned-aerialsurveys(NeiningerandHacker,2011).These

systemshaveevolvedrapidlyoverthepastdecade,aprocessdrivenprimarilybytheirvarioususesinmilitary

operations. Only recently have they become available for civilian and scientific uses, for instance for earth

sensingreconnaissanceandscientificdatacollection(Wattsetal.,2010).Recentdevelopmentsthathaveturned

present-dayUAS into realisticalternatives tomanned systemsare longer flightdurations, improvedmission

safety,flightrepeatabilityduetoimprovedautopilotsystems,andreducedoperationalcostswhencompared

tomannedaircraft.Advances inquadcoptersystemshaveproventoreducetheneedforextensiveoperator

trainingandoperatingdifficultiescomparedwithfixed-wingaircraft,particularlyinflightswithshortrangesthat

mayrequirelowerspeedsandaltitudes.Theactualadvantagesofanunmannedplatform,however,dependon

manyfactors,suchasaircraftflightcapabilities,sensortypes,missionobjectives,andthecurrentUASregulatory

requirementsforoperationsoftheparticularplatform(outlinedinsection9.4.1)(Wattsetal.,2010).

UAShavebeenwidelyusedfordifferentpurposesbothforresearchandindustrysurveys.Someofthework

developedinresearchhasincludedstudiessuchasmeteorologicaldata(e.g.FunakiandHirasawa,2008),sea

icemonitoring(e.g.Inoueetal.,2008),andwildlifemonitoring(e.g.Hodgsonetal.,2010).Shellhasconducted

a few studies in offshore Arctic regions in the last decade, some showing the full potential of this type of

equipmentforindustrialoperations(Koskietal.,2009a;Lyonsetal.,2006).Additionally,differentstudieshave

beenconductedtotestindividualsystemsandsensorsinawaytobetterunderstandtheircapabilities(Table7).

Inoffshoreareas,therehasbeenanintensefocusonseaicemonitoring,pipelineinspectionandwildlifesurveys

withUAS,whichreducestherisktohumanpersonnelandarecompletedatafractionofthecostcomparedto

theuseofmannedaerialplatforms.TheavailabilityofcommercialandsmallUAShasalsopromptedtheuseof

thistypeofequipmentinresearch,withoutrequiringlargebudgetsorinstructedpersonnel.However,withthe

continuousdevelopmentofplatformsandsensorsthecapabilitiesofthesesystemstomonitoroffshoreregions

anddetectmarineorganismsevolve.

10Dull as marine mammal surveys can be considered dull – spending hours looking for an animal either from plane or ship, most of the time not seeing anything. Dirty as fieldwork can be dirty – sea sickness or plane sickness, few clothes, maybe no access to shower and clean toilet conditions, bad weather and sea salt can affect equipment and log sheets, etc. Dangerous because if there is bad weather at sea it can easily affect navigation and risk people's lives.

94


DATE:July2016


Inrecentyears,UASoperationshavebeenintegratedintoavarietyoffieldstudiesinvolvingarangeofspecies

and applications of UAS, including the use of UAS to deter birds from feeding in commercially important

agricultural areas (Grimm et al., 2012),monitoring habitat and biodiversity loss (Koh andWich, 2012), and

monitoringillegalpoachingactivities(Olivares-Mendezetal.,2013;Mulero-Pázmányetal.,2014;Smithetal.,

2016).UndertheMarineMammalProtectionAct(MMPA),NOAA'sNationalMarineFisheriesService(NMFS)is

challengedwithdevelopingfurtherresearchandstockassessmentsforbothcetaceansandpinnipeds,andhas

beendevelopingaUASprogramto improvescientificapplications.Mostofthepublishedresearchprograms

andactivitiesareontheapplications,capabilities,andinventoryofUAStechnology.However,thereremainsa

missing component on the application of these systemswith amore industrial approach focusing on their

benefits inmanagement issues and their integration with other autonomous technologies (coordination of

unmannedvehiclesaerial,surface,andunderwater,asacommunicationnetwork).

ThereareavarietyofUASsystems,includingremotelycontrolledaircraft(e.g.operatedbyapilotataground

controlstation)andaircraftthatcanflyautonomouslybasedonpre-programmedflightplansormorecomplex

dynamicautomationsystems,suchasforsearchandrescuewheretheaircraftisprogrammedtofindanobject

oritscueinaspecificarea.Theycanalsoincludevarioustetheredsystemsthatareeitherstationaryorstationary

relative to amobile anchor point. The great potential of UAS systems liemainly in their comparative cost

effectivenesscomparedtomannedaircraft,butalsostemfromthefactthatthesesystemsaregenerallysafer

andoperationallysimplertouse.Theirusefulnessformarine-basedsurveyswilldependgreatlyontheirflight

capabilities(speed,range,altitude)andontheirpayloadcapacityinrelationtosuitablesensorpackages(e.g.

cameras, thermal sensors) (Campoyetal.,2009).Payload includeseverything that isnotpartof theaircraft

structureitself,rangingfromthefuelandbatterytypetothedifferentcamerasanaircraftcancarry.Thoughit

isoftennotspecified,itisgenerallyassumedthatthepayloadassociatedwithUASrepresentssolelytheweight

oftheimagingequipment.However,thisshouldbespecifiedbythemanufacturers,assomemaydefinetheir

payload as all the components that are external to the aircraft and others may only include the imaging

equipment.

Severalstudieshavebeenconductedtotestthecapabilitiesofdifferentaerialsystems,suchaspoweredaircrafts

(e.g.Hodgsonetal.,2013;Koskietal.,2009b),kites(Fraseretal.,1999;Vorontsova,2015)andlighter-than-air

aircrafts(e.g.Flammetal.,2000;Hodgson,2007).Eachsystemmayhaveadifferentapplicability,dependingon

theweatherconditionstowhichitisexposedandthepayloadsensorsthatitcancarry.Whiletheunderstanding

oftheadvantagesandlimitationsofpoweredsystemsformarinemonitoringhasincreasedinrecentyears,there

is alsoan interest inexploring theapplicabilityof kitesand lighter-than-air aircraft, since these systemsare

quieter and therefore potentially less disturbing to animals. Furthermore, they can be considered more

environmentally friendly than fuel powered systems. However, there is much less information about the

performanceofsuchsystemsinmarineapplications,andnocomprehensiveoverviewofthedifferentsystems

andtheirapplicationtomarineorganismmonitoringhasbeencarriedoutuptodate.Theextenttowhichthese

95


DATE:July2016


systemscanprovideusefulscientificoroperationaldataiscurrentlyuncertain,andtheuseofUASformonitoring

andmitigationpurposesinrelationtomarineorganismsisthereforeverymuchatanearlyteststage.

ThereisawiderangeofdifferentUASwithsizesassmallas18grams(PD-100byProxdynamics)to14,000kg

(GlobalHawkbyNorthropGrumman).However,thecomplexityofthelargersystemsrequiresalotofpersonnel

to operate andmany of the systems are only available formilitary use. Larger platformsmay also require

runwayssimilartothoserequiredbymanned-airplanes,whichinturnlimitsthepossibilityfordeploymentsat

sea.Smallerplatforms,ontheotherhand,mayhaveastrongersensitivitytoenvironmentalconditions,andare

limitedinthepayloadweighttheycancarry.Tetheredsystemsmayovercomesomeoftheseproblems,butin

turnarelimitedintheiroperationalrange.

9.2.1.2 AutonomousUnderwaterandSurfaceVehicles

AUVandASVareunmanned,self-propelledvehiclesthatcanoperateindependentlyforperiodsofafewhours

toseveralmonthsorevenyears.AwidevarietyofAUVandASVarenowavailableforpurchaseandhire.These

range from relatively small vehicles,which can be lifted by one or twopersons and deployed froma small

inflatable,tolargediesel-poweredsurfacevesselswithbespokelaunchandrecoverysystems(Griffiths,2002).

Thesmallervesselsareoftenoperatedwithahighlevelofautonomyandcanstayatseaforseveralmonthsat

atime.Thelargersurfacevehiclestendtobemorecloselycontrolled,althoughthisrestrictionismostoftenfor

regulatoryandsafetyreasonsratherthanafundamentaloperationalconstraint.

Many AUV’s and ASV’s were initially developed for military applications (e.g. mine sweeping, battle space

preparation)andlongtermacademicsurveys.TheyhavealsobeenextensivelyusedinO&Goperationssuchas

forsubseaequipmentinspections,leakdetection,dynamicpositioning,etc.Itistimelytoconsidertheuseof

AUVandASV in theenvironmentalcommercialsector,specificallyoilandgasoperations.Key issuesarethe

vehicles’deploymentduration,healthandsafetyconcernsassociatedwiththehazardsofrefuellingatsea,and

deployment, recovery and control of the vehicles in adverse conditions. Autonomous vehicles may have

particularadvantagesinareaswheretheycanbeeasilydeployedandrecoveredfromland.However,forlong-

termdeploymentatseainvolvingmaintenancerefuelling,morefrequentdatacollectionandsupportstaffmay

berequired,introducingarangeofcomplexlogisticalproblems.Additionally,deploymentriskismoresignificant

inbusyareasofhighmilitary, shippingor fishingactivity,due toacoustic interference,collision riskandnet

entanglement (Wynnetal.,2013).However, thepotentialofAUVandASV is clearlygiven in theirability to

acquiredataininaccessiblepartsoftheoceanatshortnotice(giventhathiringvesselsorsurvey-planesmay

require long lead times) and provide improved temporal and spatial resolution of a broad range ofmarine

measurements.

Inthenavalsector,nicheapplicationshavebeenidentifiedtogivecleardevelopmentalguidanceoncraftand

sensortyperequirements.Intheresearchworld,applicationaimsremainmuchmoreopen.Scoresofplatforms

(particularly ASV) exist in the developmental stages and trialling tends to remain geared toward system

96


DATE:July2016


improvementasageneralaiminitself.Withintheindustrysector,veryfewcraft(particularlyASV)appeartobe

genuinelyavailablewithcommerciallyviableapplicationsstillbeingestablished.Incomparison,AUVareslightly

moreadvancedcommercially.Thoughthereisclearpotential,thecombinationofcraftandsensor(s)thatmeets

anicheiscurrentlyinthepioneerstage.ThepresenceofASVinthemarketiscertainlygrowing,especiallyas

theyareincreasinglyrecognisedasavalidalternativetoAUVinmanyapplications.RelativetoAUVhowever,

regulatory uncertainty are a constraint to the wider use of ASV (see section 9.4.2 for further details). The

relationshipbetweenASVandAUVisalsocomplementaryhoweverandthereisgrowinginterestintheuseof

ASVtosupportAUVinnavigationandasadatarelay/transmissionstation.

Many surface and underwater autonomous vehicles are used on a regular basis for the collection of

oceanographic data by research institutions world-wide (e.g. Figure 11). Primarily, they are used for the

collectionofoceanographicdataand thedetectionofhigherorganisms, suchasmarinemammals and fish,

generallyremainsanicheareaofresearch.SeveralpublishedpapersdescribefieldtrialsofAUVandASVfor

environmentalandoceanographicapplications(summarisedinTable8),butmostofthepublishedworkfocuses

ontechnicalaspectsofautomatednavigationandcollaborativeperformanceofAUV.

Figure11.TheUKNationalOceanographyCentresAutonomousVehicleFleet11

AUVandASVarecapableofcarryingavarietyofsensors fore.g.seabedmappingandacoustic relateddata

collection,including,butnotlimitedto,Multi-BeamEchoSounders,Sub-BottomProfilers(SBP),side-scansonars

(SSS),magnetometers, geochemical instruments, imaging systems (HDcameras),oceanographic instruments

(Conductivity-Temperature-Depthunits(CTD),echosounders,AcousticDopplerCurrentProfilers(ADCP),water

11 http://www.seebyte.com/wp-content/uploads/2014/09/NOC_019.jpg

97


DATE:July2016


samplers and passive acoustic data acquisition systems (PAM); Wynn et al., 2013). The sensors deployed

determinethealtitude(depth),speedandenduranceofthevehicle.Forexample,higherpowersensors(SSS

andSBP)reduceendurance,highresolutionseafloorimaging(HDcamera)requireslowaltitudes(Wynnetal.,

2013).Fernandesetal., (2003) investigatedtheapplicationofAUVforfisheriesacoustics.Formarineanimal

monitoring;passiveandactiveacousticsensors(PAMandAAM)havebeensuccessfullyintegratedintoAUVand

ASV,asdumb(recordingonly)andsmart(real-timeprocessing)sensors,andreportspublishedoftheirfindings

(Table8).However,withintheexistingliterature,noreviewhasbeenfoundthatevaluatesfieldtrialsofAAMor

PAMformarinemammalsasinstalledinautonomousvehicles.

PAM systems have successfully been deployed on buoyancy gliders for the detection of baleenwhale calls

(Baumgartner et al., 2008; Baumgartner et al., 2013; Baumgartner et al., 2014b; Baumgartner, 2014;

BaumgartnerandFratantoni,2008))andforbeakedwhales(Klincketal.,2009;Klincketal.,2012;Klincketal.,

2014).Baumgartneretal.,(2014a)usedaDMONdetectormountedonaSlocumElectricGlider,programmedto

detectthetonalvocalisationsofseveralbaleenwhalespecies.TrialswereconductedofaDecimusPAMsystem

onanSV2wavegliderinthespring2014offthecoastofScotland,successfullydetectingharbourporpoiseclicks,

dolphinwhistles,spermwhalesandseismicairgunnoise(D.Gillespie,pers.Comm.).Harbourporpoiseswere

alsodetectedusingamodifiedDMONsystemonasubmarinegliderofftheSWoftheUK(Subergetal.,2014).

The British based companyMOST trialled their own wave-powered vehicle (Autonaut) in 2014, which was

equippedwithashortstreamerarray,althoughunfortunatelythishadinsufficientbandwidthforthedetection

ofcetaceans(D.Gillespie,pers.Comm.).

Thelevelofreal-timereportingfromtheseinstallationsvariedconsiderably.Baumgartneretal.,(2013)were

abletosendsufficientdatatoshoretodetectthepresenceofbaleenwhales innearrealtime.Klincketal.,

(2012)sentsummarydataforeachdivetoshore,butahighfalsealarmratemeantthatitwouldhavebeen

difficulttousethesedataforreal-timedecisionmaking.ThesystemdescribedinSubergetal.,(2014)hadno

real-timereportingcapability. Inmostcases,onlyasinglehydrophonewasdeployedoneachdevice,so the

vehicleswereunabletoprovidelocalisationinformation.TheexceptiontothisisastudydescribedbyFucileet

al., (2006),whoreportonadeploymentwhere threeSlocumgliderswereequippedwith timesynchronised

hydrophones.Thegliderswereprogrammedtomaintain stationapproximately5 to7kmapartand timeof

arrivaldifferenceswereusedtolocaliseanimals.PAMsystemshavealsobeensuccessfullydeployedfromboth

profilingandsurfacefloats.Walletal.,(2012;2014)describestudiesinwhicharecordingdevicemountedona

SlocumgliderwasusedtodetectfishsoundsoffWest-centralFlorida.Severalspeciesofsoniferousfishwere

detectedaswellassomeunknownsoundsbelievedtobeofbiologicalorigin.Matsumotoetal.,(2013)describe

aPAMsystemdesignedtodetectbeakedwhaleswhichusesthesameacoustichardwareasdescribedbyKlinck

etal., (2012)attached toanAPEXprofiling float12. Lowcost surfacedriftersarealsounderdevelopmentby

12 http://www.webbresearch.com/apex.aspx

98


DATE:July2016


researchersatNOAAfisheries(Griffiths,2015).Duetotheirrelativelowcost,profilingfloatsanddriftershave

thepotentialtobedeployedinlargenumbers,howevertotallackofnavigationalcontrolmaylimittheiruseto

regionswithspecificcurrentregimes.PAMsystemshavealsobeenincorporatedintoanumberofsmallsurface

vehiclesincludingtheLiquidRoboticsWaveglider,theASVC-EnduroandtheMOSTAutonaut,howevernopeer

reviewedliteratureyetexistsdescribingthesedeployments.

We are not aware of any studies which have collected PAM data from powered underwater vehicles. The

expense,shortmissiondurationandnoiseoutputfromthesevehicleswhichcanbothmaskPAMdetectionsand

affect the behaviour of marine organisms has led the research community to show little interest in these

platformsforPAM.

AcousticreceiversforthedetectionofacoustictagshavebeenintegratedintobothunderwaterglidersandAUV

(Eileretal.,2013;Haulseeetal.,2015).Therearealsoanumberofarticlesonlinethatdiscusstheintegrationof

acoustictagsfortrackingfishandothermarinelife131415.

AAM have already been integrated into underwater gliders and ASV (Table 8) and have focused on using

echosounderstodetectzooplanktonratherthanfishandotherlargerorganisms,althoughwehaveincluded

theminthisreviewasproofofconcept.AAMisusedfrequentlyinfisheriesresearchtoprovidefishbiomassand

distributionestimates (e.g. herring, bluewhiting,mackerel, sardines andanchovy;Doray, 2012;Huseet al.,

2015).For thepurposesof this review,wehave identifiedthe followingchallenges for their integration into

autonomousvehicles:

(1) Transducersizebalancedwithbeamwidthbalancedwithrequiredfrequency.Forexample,onlyhigh

frequency(120kHz)widebeamangle(10degrees)areusedonsmallautonomousplatformssuchas

ImagenexES853onSeaglider(Guihenetal.,2014),

(2) Powerconsumption(mosthullmountedtransducerstransmitat0.2to2kWatt),and

(3) Datatransmissionforreal-timedecisions(eventhesimplesingle-beamechosounderrawdataistoo

largetotransmitbackoverIridium).Currentlyacousticprocessingisnotundertakenonboard,butthis

couldbeanareaforfuturedevelopment.

AAMhavealsobeenusedtodetectseveraldifferentspeciesofmarinemammal,however,thesestudieshave

notbeenconductedwithAUVorASV,butratherwithAAMsensorsdeployedfromvessels,fixedstructuresor

intanks.WeincludeexamplesofthisworkheretodemonstratethepotentialoftheAAMsystemsformarine

mammal(andotherlargeanimal)detection:greywhales(LucifrediandStein,2007),finwhales(Nøttestadet

al.,2002),humpbackwhales(Love,1973),bowheadwhales(Pycetal.,2015),rightwhales(MillerandPotter,

13 http://www.futurity.org/glider-fleet-to-track-fish-in-real-time/ 14 http://news.uaf.edu/underwater-gliders-may-change-how-scientists-track-fish/ 15 http://www.gizmag.com/wave-glider-sharks-stanford/23758/

99


DATE:July2016


2001),bottlenosedolphins(Au,1996),duskydolphins(Bernasconietal.,2011),spinnerdolphins(Benoit-Bird

and Au, 2003b; Benoit-Bird and Au, 2003a), killer whales (Knudsen et al., 2008), West Indian manatees

(Gonzalez-Socoloskeetal.,2009;Gonzalez-SocoloskeandOlivera-Gomez,2012),greyseals(Hastieetal.,2014),

harbour porpoise (Hastie, 2012). A wide range of different AAM sensors have been used in these studies

including:theSimradSA950,EK60andEK500,theHumminbird987cSI,theTournamentMasterFishfinderNCC

5300,theKongsbergSM2000,theCodaOctopusEchoscope2andtheTritechGemini.AAMhasalsobeenused

to detect shark species including bull sharks, greatwhite sharks and tiger sharks (Parsons et al., 2014) and

baskingsharks(Lieberetal.,2014)usingtheTritechGeminiandtheReson7128sensorsystems.Therearealso

afewexampleswhereAAMdeviceshavebeenusedtodetectandinvestigatethetargetstrengthofvarious

turtlespecies(Mahfurdzetal.,2015;Perez-Arjonaetal.,2013;MahfurdzandHamzah,2014;Mahfurdzetal.,

2013;DavyandFenton,2013).

9.2.1.2.1 Autonomousunderwatervehicles

AUVhavebeenprimarilyappliedinseabedimagingfortopographiccharacterisationandhabitatassessmentof

benthicspecies(e.g.Williamsetal.,2010),environmentalmonitoring,hydrographicsurveys,anddetectionand

density estimation of zooplankton (Brierley et al., 2002) and fish schools (Fernandes and Brierley, 1999).

Fernandes and Brierley have also carried out studies on the avoidance response of krill and fish to vessels

(Brierleyetal.,2003;Fernandesetal.,2000).

Inthecommercialsector,sitesurveyingandgeo-hazardassessmenthavebeentheprimaryuseofAUVs.Other

typical commercial applications of AUV are pipeline monitoring, oceanographic surveys and water quality

assessment, geophysical surveying and coastalmapping. Environmental and oceanographic applications are

beginningtocrossoverintothecommercialsector,inwhichASVhaveprimarilybeenutilisedinbathymetricand

environmentalsurveying(Cacciaetal.,2009).Semi-submersiblemodelshavebeenusedforbathymetricand

hydrographic surveys inmilitary, civil and commercial applications (Wolking, 2011). Self-powered crafthave

beenusedformeteorologicalandoceanographicdatacollectionaswellassomeenvironmentalmonitoringin

boththeresearchandcommercialsectors(Cacciaetal.,2009).

For naval purposes, AUV have been utilised formine sweeping, battle space preparation, harbour security,

submarine detection and ISR (Intelligence, Surveillance and Reconnaissance). They have been deployed for

harboursecurity,minesweepingandalsousedasordnancetargets.Theseapplicationshavefrequentlybeen

basedonremotelyoperatedsurfacecrafts(e.g.ASV’sC-WorkerandTargetfamilies).Militaryinterestforself-

poweredcraftisunknown,buttherewouldappeartobescopeforasurveillanceapplication

ForAUV,asGPSreceptionislostduetotheantennabeingsubmerged,alternativesolutionsfornavigationare

requiredforpositioningunderwater.AUVnavigateusing1)arraysofacousticbeacons(transponders)onthe

seafloor for Long Baseline (LBL) communication, or 2) a combination of Ultra Short Baseline (USBL)

100


DATE:July2016


communications,GPSpositioningandinertialnavigation,applyingdeadreckoningwhenbelowthesurfaceusing

combinedinformationfromdepth,inertialandDopplervelocitysensors(Wynnetal.,2013).

Propelled AUV originate from the modern torpedo, a self-propelled projectile driven by compressed air,

inventedbyRobertWhiteheadin1866.Whitehead’storpedotravelledataspeedof3m/s,coveringdistances

ofupto700m.ThefirstAUVwastheSpecialPurposeUnderwaterResearchVehicleorSPURV,developedby

the Applied Physics Laboratory at the University of Washington in 1957, used to study diffusion, acoustic

transmissionandsubmarinewakes.Workdoneinthe1960scomprisedtheinitialinvestigationsonpossibleAUV

applications. In the 1970s there were important advances in the development of the technology, and the

experimentationwithprototypesnoticeablyincreasedoveracoupleofdecades.Inthe1990s,experimentation

movedintoafirstgenerationofAUVaimedtoperformdefinedtasks.TheproliferationofAUVmodelsledto

theexistenceofaround75AUVworldwidein2001(Fernandesetal.,2003),whenthefirsttrulycommercial

productsbecameavailable(Blidberg,2001).

Autonomousunderwaterbuoyancyglidersarerevolutionisingthewaythemarineenvironmentisinvestigated,

monitoredandassessed(Stommel,1989;Davisetal.,2002;Griffithsetal.,2007).Theyenableobservationsto

bemadeinregionspreviouslyinaccessibletovessel-basedinstrumentse.g.beneathpolaricesheets(Sagenet

al.,2008;Sagenetal.,2009;Jonesetal.,2011a),orinhurricanesandtropicalstorms(Milesetal.,2015),and

continue to operate in weather conditions (both visibility and sea state) that would limit vessel-based

observations (Milesetal.,2013).Theyprovideextremelyhighverticaland lateral resolutiondata thatmake

observations of oceanmechanisms at scales that are not resolved using conventional measurements from

vessels(Heywoodetal.,2014;Kaufmanetal.,2014).Theyarearelativelyinexpensivetechnology,withthecost

oftheplatformanditsdailyrunningtypicallyordersofmagnitudelessthanthatofvessels(Davisetal.,2002).

Frequentlythisenablesmultipleplatformstobedeployedsimultaneously,significantlywideningthespatialand

temporalresolutionofthesurvey(Rudnicketal.,2004).Comparedwithmanyothervesselstheyareacoustically

quiet(WoodandMierzwa,2013),makingthemanidealplatformtomeasurebackgroundnoise(Baumgartner

etal.,2014b).Finally,whilstfrequentlyusedforopenoceanresearch,theyaretypicallysmall,lightweight,and

deployablebyasmallrigidhulledinflatableboat(RHIBor“RIB”)withacrewofonlytwopeoplemakingthem

highly versatile (Davis et al., 2002). In the last ten years gliders have moved from being experimental,

developmentalplatforms,toformingkeycomponentsofroutinecoastalobservationalnetworks(Schofieldet

al.,2007;Schofieldetal.,2010).Theirapplicationindefence,industryandpolicysectorsisalsobeingexamined

(Wynnetal.,2013).

Theabilitytoreceiveinstructionstoadjustoperationalparametersandtransmitdatasetsquicklyisakeyfeature

ofunderwaterbuoyancygliders.Nearlyallglidersuselow-earthorbitsatellitecommunications(andinparticular

theIridiumsatellitephonesystemforbi-directionalworldwidecommunication),severalalsohavethecapability

touseradiofrequency(RF)LocalAreaNetwork(lineofsightRFmodem),CircuitSwitchedCellularandacoustic

101


DATE:July2016


modem communication. ARGOS16 is commonly used as an emergency back-up locator. ARGOS transceiver

modulesaremountedonnear-polarlow-orbittingsatellites.Atthepoles,eachsatellitepassesapproximately

twice a day (equal to 28passes), at the equator there are 6 to 7 passes in total. Location estimates for an

instrumentrequireaminimumoftworeceivedmessagesduringasinglesatellitepass.Duetothelownumber

ofsatellitepassesatlowlatitudes,gapsofafewhourscanoccurwherepositioninformationisnotpossible.All

systems,exceptacousticmodems,aresubjecttolossofperformanceinhighsea-states(Davisetal.,2002).The

amountofdataandcommunicationundertaken isabalancebetweenthebandwidthof thecommunication

method,thecostofdatatransferand,withtheexceptionofacousticmodems,thetimetaken.Whenatthe

surfaceagliderisnotincontrolofitsnavigationandthereforebothsusceptibletooceancurrentsandcollisions

withothervessels.

Eitherwaypointsorheadingdirectionareusedtodirecttheunderwaterglidermissionandglidersaresenton

transects(e.g.Piterbargetal.,2014)ortoactasavirtualmooring,profilingupanddownatasinglelocation

(e.g.Smeedetal.,2013).Whencurrentspeedsarelessthanthegliderspeed,aglidercanperformrepeated

profileswhilemaintainingatanearlyconstanthorizontalposition(Rudnicketal.,2004).Theprimarynavigation

systemofunderwaterglidersusesanon-boardGPSreceivercoupledwithanaltitudesensor,adepthsensor

andanaltimetertoprovidedead-reckonednavigation.Asimplenavigationalgorithmisusedtocomputethe

headinganddiveanglerequiredtoreachadesiredlocationfromthecurrentposition.Themajorityofglidersdo

not compensate forwater currents; the Seaglider is unique in its ability to estimatewater currents using a

Kalmanfilter(Benderetal.,2008).Inregionswheresurfaceaccessischallenging(asaresultoficeorheavyship

traffic), alternative means of location are acoustic base-line or geophysical-aided navigation methods or a

surfacevesselshadowingtheAUV(ClausandBachmayer,2015).TheuseofRAFOS(RangingandFixingofSound),

typicallyanetworkofsea-floormountedbaselinetranspondersasreferencepointsfornavigation(anacoustic

base-line),hasbeensuccessfullyimplementedintogliders(Jonesetal.,2011b;DAMOCLES).Thegeophysical-

aidednavigationmethodsproposedbyClausandBachmayer,(2015)involveterrain-aidednavigationusingthe

glidersdead-reckoningnavigationsolution,theon-boardaltimeterandalocalDigitalElevationModel(DEM-

3Drepresentationofaterrainssurfacecreatedfrombathymetricdata).

Themajorityofbuoyancyglidersarecurrentlybatterypowered,andcarryanincreasingvarietyofsensorsof

interesttooceanographers(Rudnicketal.,2004).Theseincludephysicalvariables(e.g.pressure,temperature,

salinity,currents),noise(e.g.background,ambient,ships,marinemammalcalls,fish),chemicalvariables(e.g.

dissolvedoxygen,CDOM,nitrate,pH)andbiologicalrelevantvariablessuchasabundanceofphytoplanktonand

zooplankton(Rudnicketal.,2004;Davisetal.,2008;Guihenetal.,2014;Klincketal.,2012).Threetypesof

sensorscanbefittedtoglidersthathavethepotentialtomonitormarinemammals,turtlesandfish:passive

acousticmonitoring(PAM)sensors,activeacousticmonitoring(AAM)sensorsandanimaltags.

16 www.argos-system.org

102


DATE:July2016


9.2.1.2.2 Autonomoussurfacevehicles

CivilianapplicationsofASVconsistbasicallyinbathymetricandenvironmentalsurveyandmonitoring(Caccia

etal.,2009).Militaryapplicationsincludeharboursecurity,coastalsurveillance,minesweepingandsubmarine

detection.ASVhaveundergonethemostappliedtriallinginthedefencesector.ThedevelopmentofASVwasin

apioneerstageby2006andtherearenowmanyoptionstochoosefrom(Caccia,2006;Cacciaetal.,2009).

PoweredASVresearchfocusesonmeteorologicalandoceanographicdatacollection,withbathymetryimaging

beingthemostcommonsurveyingactivity(e.g.Vanecketal.,1996).Otherimportantenvironmentallyoriented

studieshavebeenfocusedonautomatedfishtracking(Goudeyetal.,1998)andacquisitionofwatersamplesin

theseasurfacetopmicrolayerfortheanalysisofclimaticchanges(Cacciaetal.,2005).

Theautomatednavigation technologyused inmodern civilianASV,bothpoweredand self-powered,has its

origininthreeautonomousvesselscreatedbyMITduringthenineties(ARTEMIS,ACESandAutoCat)andthe

autonomous kayak SCOUT. These demonstrated the feasibility of automatic heading control, way-point

navigationbasedonDGPSandfullyautomatedcollectionofhydrographicdata(Cacciaetal.,2009).Powercan

bederivedfromavarietyofsources,includingdieselmotors,fuelcellsandbatterypacks.Electricalpowersupply

isgenerallypreferredinenvironmentalsamplingapplications,whichrequirenopollutionoftheoperatingarea;

diesel propulsion is preferred in long missions (e.g. coastal surveillance or Mine Counter Measure (MCM)

operations).Thesevehiclescantravelathighspeedsandhavethecapacitytocarry/towsignificantpayloads.

9.2.1.3 MissionPlanningandVehicleControl

Forsafetyandlegalreasons(seesection9.4)aircraftandlargesurfacevehiclesaregenerallynotoperatedout

oflineofsightofahuman“pilot”,althoughthereisnofundamentalreasonastowhytheycouldnotbegiven

greaterlevelsofautonomyiflegalandsafetyconstraintscanbeovercome.AdvancesinGPStechnology,and

nowAIS,havebeenkeytotherapiddevelopmentofpoweredaswellasself-poweredASVformanyresearch,

commercialandmilitaryapplications.ASVnavigationatthewatersurfaceallowsthemtorelayhighfrequency

transmissionsintheairandacoustictransmissionsinthewater(ASVcanbeusedtosupportAUVnavigation).

ASVareoutfittedwithGPSreceiversandattitudeandheadingreferencesystem(AHRS)sensors.AUV’sonthe

otherhand,canonlygatheraccurateGPSpositionswhentheycometothesurface.Otherancillarysensorsare

Dopplervelocitylogs(DVL),whichmeasurethefluidflow(currents)andallowforbettercontroloftrajectories.

Typically, in an autonomous vehicle, lower level actions (like rudder control) are automated and overall

behaviour(likewaypointselection)ismanagedbyanoperator.Fullautonomyisdesiredinresearchandcivil

applications;inmilitaryapplicationsremote-controlledvesselsarepreferred,andautomationisregardedasa

waytooptimisesystemperformance(Alves,2002;Caccia,2006).Variousoptionsofcontrolinclude:

• Controlinthehorizontalplane:fromsimpleheadingcontroltomorecomplextechniques.

• Trajectory tracking: Capability of the vessel to follow a time-parametrised reference curve in two-

dimensions(controlofthepositionofthevesselatspecificinstants),whichisespeciallychallengingin

103


DATE:July2016


presenceofexternaldisturbance(waves,wind,currents).Timeconstraintsaremorerelaxedinpractical

applications.

• Path following: Capability of the vessel to follow a reference curve in two-dimensions, without

temporalconstraints.Adesiredtemporalspeedprofilewillstillbeapplied.

• Cooperative motion control: Capability of the vessel to adapt its trajectory in the presence of

automatedmarinevehicles. Interestingapplicationsare theuseofanASVasa communication link

betweenanAUVandasupportvessel(leader-followerproblem)orthecoordinationofmultipleASV

forhydrographicsurveying.

• Missioncontrol:AMissionControlSystem(MCS)allows theendusertoexecuteandsupervise the

progressofoneormultiplevehicles.Missionplanningandcontrol is limitedtothedefinitionofthe

trajectorybymultiplepointsandafewemergencyactions(e.g.abortandstop).

104


DATE:July2016


Table7:ListofpublishedstudiesinvolvingUASwithapplicationinthemarineenvironment.

Reference Objective UASType Sensor Datatype Studyconclusions Country

Bevanetal.,

2015

Evaluatethe

effectivenessofalow-

costcommercially

availableUASfor

identifyingbothadultand

hatchlingseaturtlesin

near-shorewaters

adjacenttonesting

beaches.

DJIPhantom

1

quadcopter

GoProand

GPSenabler

Video TheresultsindicatethatthisUASsystemprovidesapracticalandeffective

methodofconductingdaytimesurveysinnear-shorewatersfor

monitoringseaturtleabundanceandmovements.Eveninturbidnear-

shorewaters,theywereabletomonitorthesurfacingofadultfemalesas

theyleftthenestingbeach.WhiletheUASwascontrolledfromshorein

thecurrentstudy,itcouldalsobelaunchedandcontrolledfromaboatfor

monitoringturtlesinoffshoreareassuchasforaginggrounds.

Mexico

Cameronet

al.,2009

NOAAtestsofaUASto

determineits

effectivenessfor

surveyingsubarcticpack

iceforiceseals.

ScanEagle

aircrafts

SingleLens

Reflex(SLR)

NikonD300

andvideo

Photoand

video

TheScanEagleperformedwellinavarietyofweatherconditions,andthe

imagescollectedhavethenecessaryresolutiontodistinguishthedifferent

species,ages,andoccasionallyeventhegenderoficeseals.

Beringsea

(USAand

Russia)

Durbanet

al.,2015

UASasanalternative

methodforsuccessfully

obtaining

photogrammetryimages

ofkillerwhalesatsea.

APH-22

hexacopter

(Aerial

Imaging

Solutions)

OlympusE-

PM2,

Olympus

M-Zuiko

25mmf1.8

lens

Videoand

photo

(triggered

manually

with12.3

MP)

TheAPH-22hexacopterhasgreatutilityforcollectingphotogrammetry

imagestofillscientificdatagapsaboutfreerangingwhales.Itissmalland

portablewithVTOLcapabilitythatenablessafedeploymentandretrieval

evenfromsmallboatplatforms,andenablesphotogrammetryinremote

locationswhereconventionalaircraftsareimpractical.Canbeflowninlow

altitudeswithoutdisturbingthewhales.Allowsdifferentiationofindividual

whalesusingnaturalmarkings,withprecisealtitudetoenablequantitative

measurements.

Canada

Hodgsonet

al.,2010

CoastalUAStestsand

mannedvsUAS-

differenttests.

Powered

Warrigul

SonyDigital

Recorder

Video ThecombinationofthetypicalUASimagingsystemweusedandthe

altitudestrieddidnotprovideimagesofhighenoughresolutiontoreliably

detectdugongsorwhales.

Australia

105


DATE:July2016



Hodgsonet

al.,2013

Coastalmonitoringof

DugongsusingUAS

Powered-

ScanEagle

SLRcamera

Nikon®D90

12

megapixel

Photo TheScanEaglewouldneedtofly2.8timesasmanytransectstoachieve

thesameareacoveragethatastandardmannedsurveycouldachieve.

Thislimitationcouldbeaddressedbyusingacamerathatcapturesimages

atahigherresolutionsothatonecoulduseawiderlensorflyhigher,or

onecouldusemultiplecameras.

Australia

Jonesetal.,

2006

Landandcoastal

assessmentofanUASfor

wildliferesearch.

MLBFoldBat CanonElura

2

Video AlthoughtheFoldBatUASsystemdidnotproveusefulasanoperational

wildliferesearchtool,ithadanumberofcharacteristicsthatilluminated

systemrequirementsforsucceedinggenerationsofUASdesignandmore

successfulUASmissions.TheElura2producedsatisfactoryimagesfor

wildliferesearchapplicationstargetingsmallandinconspicuousspecies

suchassmall,whitewadingbirds.However,theutilityoftheimageswas

limitedbecausetheywerenotinstantaneouslygeoreferenced.

USA

Koskietal.,

2009b

Comparisonoftraditional

offshoresurveywith

humansurveyorsvsaerial

digitalimagemonitoring.

Powered-

InsightA-20

Alticam400 Video UAShasthepotentialtoreplacemannedaircraftduringsurveysforlarge

cetaceansorlargegroupsofsmallcetaceansifthesearchareaissmall.

However,highervideoresolutionisneededbeforetheUASwouldbe

effectiveforsurveysoflargeareasorfordetectionofsmallercetaceans

andpinnipeds.

USA

Koskietal.,

2013

Comparisonoftraditional

offshoresurveywith

humansurveyorsvsaerial

digitalimagemonitoring.

Fixed-wing

manned

with

cameras

NikonD800

DSLR,video

CanonHD

XF100

HDvideo

and

DigitalSLR

photos

Theanalysessuggestthatimageryfromthesecamerasprovidessimilar

qualitydatatothosecollectedbyobserversflyinginanairplane,andthat

UASsurveyscanbeflowninsomeconditionswhenmannedaircraft

cannot,itappearsthatUAScouldreplacemannedaerialsurveysfor

marinemammals.However,thesamplesizesfortheseanalyseswerelow

anddidnotcoveralloftheconditionsthatwouldbeencounteredduring

mannedaerialsurveyssoadditionaltestswithlargersamplesizesand

coveringawiderarrayofconditionsarerecommended.

USA

106


DATE:July2016



Thamm,

2011

Equipmentevaluation. KiteSUSI62 Not

specified

TheSUSI62wasdesignedasarobustandsafeUASwithalargepayload

andlongflighttimefordaytodayapplications.Itscapabilitytooperate

differentsensors(e.g.optical,multispectralandthermal)atthesame

timeandtheeaseofswappingsensorsoffersfascinatingoptionsfor

researchandcommercialapplications.Theautopilotguaranteesthatthe

areaofinvestigationiscoveredcompletelywiththedesiredoverlapand

groundresolution.

Europe,

Africa

107


DATE:July2016


Table8.ListofstudiesinvolvingAUVandASVwithpossibleapplicationinthemarineenvironment

Reference Objective AUV/ASVtype Sensor Datatype Summary Country

Ainleyetal.,

2015

Presence/

absence

Seaglider Imagenex

ES853

AAM AnechosoundermountedontheSeagliderwasusedtodeterminethedepth,

distributionandabundanceofprey(assumedtobekrillandfish).Thegliderwas

deployedfor78daysmakingobservationsofthephysicalandbiological

environment.ThepaperappliedthemethodofGuihenetal.2014.

Antarctica

Armstronget

al.,2006

Characterisethe

coralreefhabitat

Propellertwin-

hullAUV

Seabed

PixelflyCCD

camera

Image TheAUVSeabedequippedwithaCCDcameracapturedimagesevery2.5softhe

benthiccommunitiesthatinhabittheshelfcoralreefoftheHindBankmarine

conservationdistrict.Thedatawasusedtoprovideinformationonbenthicspecies

compositionandabundance.

St.Thomas

(USVirgin

Islands)

Baumgartner

etal.,2008

Acoustic

recordingsfrom

autonomous

platforms.

Slocumcoastal

gliders

custom

built

PAM

Obtainedgoodqualityacousticrecordingswithnodetectableflownoisewhile

profilingupordown.Glidernoisewasdetectableatthetopandbottomofeach

divepreventingdetectionofwhalevocalisationsatthesametime.

GreatSouth

Channel

Baumgartner

etal.,2013

Presence/

Absenceof

baleenwhales

Slocum DMON PAM Ahardwareandsoftwaresystemwasdevelopedtodetect,classify,andreport14

calltypesproducedby4speciesofbaleenwhalesinreal-timefromoceangliders.

Twoglidersreportedover25000acousticdetectionsandusedLFDCStoattribute

themtofin,humpback,sei,andrightwhales.Theglidersmadelotsofnoiseatthe

surface(surfacewaves,internalpumps,radiotransmissions),soDMONdetections

wereturnedoff.Detectionaccuracywashighforfinandrightwhales,modestfor

humpbackwhales,andundeterminedforseiwhales.Whalesweredetectedwith

thegliderandusedtodirecttheshiptolocatethem.

Atlantic

Baumgartner

etal.,2014b

Presence/

absence

Slocum DMON PAM ALowFrequencyDetectionandClassificationSystem(LFDCS)wasusedto

autonomouslydetectandclassifysounds,additionallyrawdatawassavedonthe

Arctic

108


DATE:July2016



DMON.Bowheadwhales,walrus,beardedseals,belugawhales,airgunsandships

noiseweredetected.Real-timereviewofpitchtracksbyanalystsonshoreallowed

unambiguousidentificationofbowheadwhalecallsandairgunpulses.Theon-

boardclassificationsystemwaslesssuccessful;thiswasattributedtotheimmature

stateofthecalllibrary.

Binghamet

al.,2012

Performanceof

theWaveGlider

foracoustic

applications

WaveGlider DMON PAM Thenoisegeneratedbythewavegliderisverylow.Theambientnoiseatthe

submergedgliderissignificantlylowerthantheambientnoiseatthesurfacefloat.

Theseastatedoesnothaveastronginfluenceonthewaveglider’semittednoise.

Southcoast

ofOahu

Binghamet

al.,2012

Performanceof

theWaveGlider

foracoustic

applications

WaveGlider ITC-3013 AAM Thewaveglidermodemwasabletoreceivesignalsfromthetransmitteratdepths

of500and2,500mandatrangesinexcessof3km.Thereliabilityandtheinput

SNRvariedwiththelocalnoisefieldaroundthesurfacereceiver.

Southcoast

ofOahu

Bourgeoiset

al.,1997

Bathymetry

mapping

Semi-

submersible

ASV

SimradEM-

1000

AAM Thesemi-submersibleASVprototypeORCA,adiesel-poweredunmannedvehicle

withatallmastforairtakeandcommunications,wasusedtoacquirebathymetric

datawithamulti-beamechosounder.Thebathymetricdataisvalidatedand

correctedinrealtime,andusedbyavessel-basedoperatortore-adjustnavigation

waypointsandoptimisesensoroperatingparameters.Thismethodologyoffers

highcoverageandaccuracyfortheacquireddata.

Florida

Brierleyand

Fernandes,

2001

Fishsurvey–the

presenceof

divingbirds

changedthe

initialobjective

PropellerAUV

Autosub-1Simrad

EK500

AAM TheAUVAutosub-1wasequippedwithanupward-lookingechosounderforfishsurveysintheNorthSea.TheechogramsfromtheAUVcontainedverticaltraces

startingfromtheseasurfacecausedbydivingbirds,identifiedasNorthernGannets

byvisualobservations.Themeandivedepthwas19.7m,somewhatdeeperthan

expectedtootherstudiedgannets.Thedivingdepthshaveimplicationsonforaging

capabilitiesofgannets,sotheeffectiveverticalforagingrangeofthespeciesshould

bereconsidered.

Shetland

andOrkney

109


DATE:July2016



Brierleyetal.,

2002

Estimatekrill

densitiesunder

Antarcticice

PropellerAUV

Autosub-2Simrad

EK500

AAM TheAUVAutosub-2wasdeployedtoestimatekrillabundanceunderAntarcticice,

uptoadistanceof27kmbeyondtheiceedge.Krilldensitiesfivetimesgreater

thaninopenwaterwereregistered.

Antarctic

Ocean

Brierleyetal.,

2003

Assessavoidance

responseofkrill

toasurvey

vessel

PropellerAUV

Autosub-2Simrad

EK500

AAM TheAUVAutosub-2wasdeployedaheadtheresearchvesselJamesClarkRossintheAntarcticOceantoassesskrillavoidancetothevessel.BothJamesClarkandAutosub-2wereequippedwiththesamescientificechosoundertoassessthe

responseofkrilltothevessel.Thesimilarityofthedensitycalculatedwiththedata

acquiredbytheAUVandthevesselsuggestedthattheavoidanceofkrilltothe

JamesClarkwasnotimportantenoughtobiastheestimationofkrillabundance.

Antarctic

Ocean

Cacciaetal.,

2005

Samplesfrom

theseasurface

micro-layer

ASVcatamaran

SESAMOMultiuse

microlayer

sampler

samplers TheASVcatamaranSESAMOwasusedintheTerraNovaBayareaoftheRossSeabetweenJanuaryandFebruary2004tocollectdataandsamplesofthesea-air

interfaceinacompletelyautomatedmanner.Thecompositionofthesea-surface

microlayer(~1mm)providesvaluableinformationaboutthephysicalandchemical

interactionsbetweenthehydrosphereandatmosphere,andtheirrelationwith

climaticandenvironmentalchanges.

RossSea

Antarctic

Sea

Correaetal.,

2012

Natureofthe

waveformcoral

ridgesofMiami

Terrace

PropellerAUV

C-Surveyor-IISimrad

EM2000

Edgetech

120kHz

AAM TheAUVC-Surveyor-IIwassentona24.5hmissiontoacquire27km2ofhigh

resolutionseabedmaps(withmulti-beamechosounderandside-scansonar),sub-

surfaceprofilesandbottomcurrentdata(ADCP)tostudythenatureofthecold-

watercoralridges,withwaveformmorphology,foundatthebaseoftheMiami

TerraceintheStraitsofFlorida.

Miami

Terrace,

Straitsof

Florida

Davisetal.,

2008

Presence/

absence,

quantity

Spray Sontek750

kHzADP

AAM SprayglidersweredeployedacrosstheSouthernCaliforniaCurrentsystemand

collectedconcurrentmeasurementsoftemperature,salinity,ADCPshear,

chlorophyllafluorescenceand750kHzacousticbackscatter.Thelongtimeseries

andhighspatialresolutiondata(28months,acrossthreelinesovertwoyears)

characterisedcloselinksbetweenfrontsinphysicalandbiologicalvariables.The

ADPwascalibratedwithtungstencarbidespheres,sovariabilityinquantitative

USA

110


DATE:July2016



measurementsis2dBbetweenmissions.Thestudyobserveddielverticalmigration

ofzooplanktonandmicronekton,butthesinglefrequencywasnotspeciesspecific

andthereforenotquantitative.

D'Spainetal.,

2004

Lowsignature

AUVforpassive

oceanacoustic

studies

Bluefin

PropellerAUV

not

specified

PAM Amid-sizeAUVisretrofittedwithavector-thrustsystemmanufacturedbyBluefin

Robotics,whichdecreasestheradiatedsoundlevelsbetween20and50dBinthe

20to10kHzband.Thelowacousticsignaturepermitstheuseofavehiclemounted

hydrophonearrayforpassiveoceanacousticstudies.

Not

specified

Fernandes

andBrierley,

1999

Testthe

capabilityofAUV

Autosub-1todetectfish

PropellerAUV

Autosub-1Simrad

EK500

AAM TheAUVAutosub-1equippedwithascientificechosounderwassentinto13missionstogatheracousticdatafordetectionanddensityestimationoffish

schools.TheabsenceofanavoidanceresponsebyafishschoolwiththeAutosub-17mfarfromitsuggestedthatfishisnotsensitivetothepresenceofthevehicle

beyondthatdistance.

NorthSea

Fernandeset

al.,2000

Assessthe

avoidance

responseoffish

toasurvey

vessel

PropellerAUV

Autosub-1Simrad

EK500

AAM Researchvesselsareusedfordensityandabundanceestimationoffish,andthere

isawidespreadconcernthatthenoisegeneratedbythesevesselsmaycausean

avoidanceresponseonfish,whichwillbiastheabundanceresults.TheAUV

Autosub-1wasdeployedaheadtheresearchvesselScotiaduringanacousticsurveyofherringintheNorthSea.TheAUVwasequippedwiththesameechosounder

modelastheScotiatoassess,bydatacomparison,theresponseoftheschoolas

thevesselgoespast.ThesimilarityofthedatafromtheAUVandthevesselleadsto

concludethatfishdonotavoidsurveyvessels.

NorthSea

Fernandeset

al.,2003

Fisheries&

plankton

acoustics

research

Autosub SIMRAD

EK500

AAM AUVcansamplepreviouslyimpenetrableenvironmentssuchastheseasurface,the

deepsea,andunder-seaice.However,advancesinpower-sourcetechnologyare

requiredtoincreasetherangeofoperationbeforeAUVaresuitableforthese

applications.

111


DATE:July2016



Fratantoniet

al.,2011

Developafully-

integrated

autonomous

system

APEXfloat,

Slocumglider,

Z-Rayglider

DMON

PAM Trialshavebeenasuccess.Thesystemislow-noise,low-power,data-efficientand

reliable.IntheSanClementeIslandtrial,boththeglidersandthefloatswereable

tosendnumerousreal-timebeakedwhaledetections.

Fucileetal.,

2006

Time-stamped

baleenwhale

vocalizations

fromaglider

SlocumElectric

Glider

HTI-96-MIN

hydrophon

ePersistor

CF2Data

Logger

PAM

Thequalityoftheacousticdatarecordedwasverygood.Theglidermovementand

flownoisedidnotaffecttherecordings.Multipleglidersdeployedatoncemeant

thatthevocalisingwhalescouldbelocalisedtowithinafewhundredmeters.

Gulfof

Maine

Goudeyetal.,

1998

Kaya-shapedASV

usedtotracka

taggedwimming

animal

KayakASV Acoustic

transducer

inatracking

tag.

AAM Asmallkayak-shapedASV,25cmlongandpoweredbyanelectricbattery,was

usedtofollowataggedswimminganimal(fish)duringanentireday(24hours).The

smallsizeandelectricpowerwereselectedtominimisethenoiseoutputofthe

vehicle.

Not

specified

Grasmueck,

2006

morphology&

oceanographic

conditions

C-Surveyor

IITMAUV

KongsbergT

MEM1002

AAM ThedataobtainedbytheAUVwasabletoprovideinformationtofillthegap

betweenlow-resolutionsurface-basedmappingandvisualobservationsonthe

seafloor.

Straitsof

Florida

Greeneetal.,

2014

Fisheries

acousticstock

assessments.

WaveGlider BioSonics

DT-X-SUB

AAM AfleetofwavegliderscouldsurveytheWestCoastUSAEEZmuchfasterandmore

frequentlythanatraditionalfisheriessurveyvessel

WestCoast

USAEEZ

Guihenetal.,

2014

Presence/

absence,

quantity

Seaglider Imagenex

ES853120

kHz

AAM Madeacousticbackscattermeasurementsthroughthedowncastandupcastofa

Seaglidermissionusingacalibratedsinglefrequency(120kHz)echosounder.

Antarctickrillswarmswereidentifiedintheacousticdataandkrillabundancein

theWeddellSeawasestimatedduringtwoshortperiodsinJanuary2012.The

studyhighlightedthecomplexityofanimaltargetorientationandgliderorientation

anditseffectontargetstrengthsandthereforebiomassestimates.Italso

Antarctica

112


DATE:July2016



highlightedtheneedforappropriatesamplingstrategiesforgliderstobeusedfor

biologicalsurveys.

Haulseeet

al.,2015

Presence/

absence

Slocum Vemco,

acoustic

receiver

AAM IntegratedVR2Cacousticreceivers(VEMCO,frequency=69kHz)intoaSlocumG2

glider.Thehydrophonesextendedoutofboththedorsalandventralhullto

increaselisteningcapabilities.Thesandtigerswerecaughtandtagged(internally)

withacoustictransmittersaspartofpreviousprojects.TheAUVdetected97%of

acoustictransmissionsfromtesttagswithinadistanceof250m.Thepercentageof

tagsdetecteddecreasedexponentiallyatdistancesgreaterthan250m.Thestudy

wasunabletodeterminetheverticalpositionofanysharkdetectedbytheAUV.

North

Atlantic

Johnson,

2012

Automatically

detectand

classify

vocalisations.

Slocumglider,

Apexprofiling

float,Drifting

surfacefloat

DMON

PAM

Beakedwhaledetectionswereexaminedtoestablishmiss-classificationratesto

improveclassificationthresholdsinthedetector.Gliderandprofilermotorswere

noisybuthadalowdutycycleandsoareconsideredviableplatforms.Profileswere

preferredwhiletheshortshallowdivesoftheglidermadeitlesseffective.

Canary

Islands

Kennishet

al.,2004

Seabedpatterns

andbenthic

habitats

PropellerAUV

Remus600kHz

side-scan

sonar

AAM Studyofthebedformsandrelatedbenthichabitatofinvertebrateanddemersal

finfishpopulationsinGreatBay(NewJersey),usinganAUV(REMUS)outfittedwithaside-scansonar.

NewJersey

Klincketal.,

2012

Presence/

absence

Seaglider HighTech

IncHTI-99-

HF

PAM TheSeagliderwasinstrumentedwithahydrophone.Acousticdatawere

compressedusingFreeLosslessAudioCodec(FLAC)andstoredonflashdrives,in

parallelacousticdatawerescreenedforbeakedwhalevocalisationsusingan

energyratiomappingalgorithm(ERMA)onboardtheglideranddetectionswere

reportedbacktoshoreonsurfacing.Beakedwhalesweredetectedon7ofthe85

dives(oneevery27.5hours),similardensitiestothoserecordedbyvisualsurvey.

Thedataconfirmedtheglideriscapableofdetectingthepresenceofbeaked

whaleswithinafewkilometers.Thehydrophonewasnotrecordingthroughoutthe

wholedive,andtheyidentifiedhowvocalizationscouldbemissedasaresultof

this.

Hawai’i

113


DATE:July2016



Lucieeretal.,

2013

Classificationof

multibeam

acousticdata

PropellerAUV

SiriusStereo

camera

system

Image TheAUVSiriuswassentonfivemissionstomapthebenthichabitatofcoastalreefs

alongtheFortescuecoastusinghighresolutionimagesobtainedfromastereo

camerasystemat1sintervals.Thecameraimageswhereusedtogeo-reference

availableMBESimages.Theresultsallowedidentificationofattributesfromthe

multibeamacousticdatawhichcouldbecorrelatedwiththepresenceofbenthic

species.

NewJersey

Manleyand

Vaneck,1998

Bathymetry

mapping

CatamaranASV

ACESBasic

recreational

depth

sounder

AAM TheASVACESisasmallcatamaranpoweredwithagasolineengineandequipped

withacontrolsystemforautomaticnavigation.Thisstudydescribestheinitialtests

ofACESasaninstrumentforhydrographicsurveys(bathymetrymapping).The72%

ofthecollecteddatametClass1standardswhencomparedtoprevioushighly

accuratebathymetrydatacollectedbytheUSACEinthesamearea.

Massachuse

tts

Matsumoto

etal.,2013

Compare

acousticfloatto

NavyM3R

system

QUEphone

acousticfloat

HTI92WB

hydrophon

eERMA

detection

algorithm

PAM

TheQUEphone’sdetectionswerecomparabletothatofM3R’s,concludingthatthe

floatiseffectiveatdetectingbeakedwhales.

AUTEC

(Bahamas)

Ohmanetal.,

2013

Presence/

absence,

quantity

Spray Sontek750

kHzADP

AAM ObservationsofzooplanktondielverticalmigrationintheCaliforniaCurrent

system.Thedeeperdaytimedepthsandlargeramplitudeofzooplanktondiel

verticalmigrationbehaviourassociatedwiththeoffshorewatersatfrontal

transitionsareattributabletoafaunisticchangeacrossthefrontandassociated

spatialchangesinlightmediatedpredationrisk.

California

Powelland

Ohman,

2015a

Presence/

absence,

quantity,

behaviour

Spray Sontek750

kHzADP

AAM SixyearsofSprayglidertransectsacrosstheCaliforniacurrentsystemare

presented.Watersonthedensersideofafrontcontainedhigherchlorophyllaand

acousticbackscatterthanwatersonthelessdenseside.Dielverticalmigration

behaviourofzooplanktonincreasedoffshoreandcovariedwithoptical

transparencyofthewatercolumn.Differencesinacousticbackscatteracrossthe

California

114


DATE:July2016



frontwerevalidatedwithnetsamplesthatshoweddifferencesinzooplankton

compositionandsizeclassesacrossthefront.

Powelland

Ohman,

2015b

Presence/

absence,

quantity,

behaviour

Spray Sontek750

kHzADP

AAM SixyearsofSprayglidertransectsacrosstheCaliforniacurrentarepresented.Mean

VolumeBackscatterStrength(MVBS)recordedfroma750kHzacousticDoppler

profilerwereusedtoobservechangesinzooplanktoncoincidentwithdensity

fronts.MVBSwassignificantlycorrelatedwithdensitygradients,anditwas

hypothesisedthatlargemobilepredatorsforaginginthevicinityofsuchfeatures

couldlocatehabitatwithhigherzooplanktonbiomassconcentrationsupto85%of

thetimebytravellinguplocaldensitygradients(i.e.,towardratherthanawayfrom

densersurfacewaters).

California

Subergetal.,

2014

Presence/

absence,

quantity

Slocum Imagenex

ES853

AAM /

PAM

TwoslocumglidersequippedwithCTDandfluorometers,onecarriedan

echosounderforresolvingzooplanktonandfish,theothercarriedamodifiedd-tag

sensor.Strongtidalcurrentsdeflectedbothgliders,whichultimatelyledto

recoveryandredeploymentofthegliders.In6weeks,thegliderscompleted5,474

divesandtravelled2,389km.Fishmarkswereobservedinthe120kHzacoustic

data,althoughnospeciesidentificationwaspossible.Harbourporpoiseclicksand

dolphinclicksandwhistlesweredetectedduringalimitedsubsectionofthe

missionbeforethed-tagpowercablewasdamaged.Noiseattributabletotheglider

pumpwasalsorecorded.

Islesof

Scilly

Vanecketal.,

1996

Bathymetrydata

acquisitionin

littoralregion

ASV(surface

craft)ARTEMISnot

specified

AAM Theprototypeautonomoussurfacecraft(ASC)ARTEMISisusedtocollectbathymetricdatainlittoralregions,forpotentialfutureuseincostalsurveysand

monitoring,minecountermeasuresandoceanographicsurveys.Accurate

geolocationofbathymetricpointsisachievedthroughaguidancecontrollerwith

differentiallycorrectedGPS.

Not

specified

115


DATE:July2016



Walletal.,

2012

Presence/

absence,

quantity

Slocum Digital

Spectrogra

mRecorder

PAM Ahydrophonewasintegratedintothecowlingofaslocumglidertomeasurefish

sounds.Redgrouperandtoadfishsoundswererecorded,aswellas3other

biologicalsoundssuspectedtobefish.Soundswererecordedalongglidertransects

toprovidebiogeographyofthesefishes.TheDigitalSpectroGramRecorder(DSG)

recordedsoundfor25severy5minutesatasamplerateof70,000Hz.Datawere

storedona16GBSDcard.TheDSGclockwassynchronisedtotheglider’s

computertostoreembeddedtimestampsintotherawdatafilestructure.Fish

soundsweremanuallydetected,asglideraltimeter,pumpandat-surfaceIridium

noisehamperedautomateddetectionmethods.

Gulfof

Mexico

Walletal.,

2014

Presence/

absence,

quantity

Slocum Digital

Spectrogra

mRecorder

PAM Threepassiveacousticglidermissionswereconductedoffwest-centralFlorida,two

inaredtide,inordertoassesswhetherredtidesinfluencesoniferousfish.In

additiontodetectionsofredgrouperandtoadfish,acuskeelsoundwasalso

identifiedandrecorded.Differentfishtypenoiseswereassociatedwithdifferent

depthlayersandwithdifferentdiurnalfrequency.

Gulfof

Mexico

Williamset

al.,2010

Studyof

drownedreef

alongtheGRB

PropellerAUV

SiriusImagenex

DeltaT

AAM Aship-basedseafloormappingsupportedbyAUVsurveyinganddredgingsamples

wascarriedoutinQueensland,tostudythedrownedcoralreefsalongtheshelf

edgeoftheGreatBarrierReef.TheAUVSiriuswasusedtogatherhigh-resolutionseafloorimageryatcloserangefromtheseabedtovalidatetheinterpretationsof

thesonardatafromtheship.

Australia

116


DATE:July2016


9.2.1 Cooperativesystems

Autonomousaerial, surfaceandunderwatervehiclesmay track individualanimalsandobtaindata fromthe

robotic platforms in a coordinated effort or a network (see Figure 12 as example). This can improve our

understandingof the individual technologies' limitations and acquire detailed information about the animal

understudyanditssurroundingenvironment.Thoughthisrequiresalotofefforttocoordinateallthesystems,

it isneverthelesshighlyvaluableforimprovingthecurrentknowledgeofindividualsystemsandhowcurrent

innovation can be applied with a broader approach to the marine environment. At the same time, this

coordinationusingglobalcommunicationanddatacentralisationallowsforagreatersituationalawarenessby

usingsatellite,radio,andinternetcommunications.InitiativessuchastheOceanObservingsystems(e.g.,ONR

Noise Observing System, NSF Ocean Observatories Initiative Coastal & Global Scale Nodes, Global Ocean

Observing System, Ocean Networks Canada, NOAA Integrated Ocean System) provide multi-scale,

multidisciplinary ocean observing systems using the interactive capability of advanced sensors and

platforms, and real-time access to data and visualization tools(NSF,2009).Thesenetworksareabletocollect

oceanand seafloordata, togetherwith informationabout theeffectsof anthropogenic activitiesonmarine

species(e.g.,Clarketal.,2008),athighsamplingratesthatcanlastyearstodecades.

Modernsurveysandobservingsystemsutilise,forexample,multiplegliderstoobservetheoceans.Technology

isbeingintegratedtocoordinateandcontrolmultipleplatformsandtotrackdynamicfeatures.Forexample:

TheGliderMonitoring,PilotingandCommunication(GLMPC)system(Jonesetal.,2011b),theAdaptiveSampling

andPrediction(ASAP)researchinitiative(Leonardetal.,2010),andtheAdaptiveAutonomousOceanSampling

Networks(AAOSN)17(seealsoZhangetal.,2007;Alvarez,2009).Thisincludestrackingtaggedorganismssuch

asfishandsharks(e.g.MASSMOproject18).Thesesystemsenvisageinterfacingautonomousdecisionmakingfor

multipleautonomousvehiclesofdifferenttypes,andarecurrentlybeingtestedinmodelandfieldscenarios19.

17 http://noc.ac.uk/sbri 18 http://projects.noc.ac.uk/exploring-ocean-fronts/robotic-vehicles-successfully-track-tagged-fish-plymouth 19 http://noc.ac.uk/SBRI

117


DATE:July2016


Figure12.SunfishTrackingproject(Pinto,2014)20

9.3 Operationalaspectstoconsider

TheoperationalaspectsoftheuseofautonomousvehiclesduringE&Pactivitiescoverthepracticalimplications

of deploying autonomous vehicles as well as trade-offs that need to be considered based on the level of

technology,abilitiesofsystemsandthecapabilitytointegratewithexistingsystemsormonitoringefforts.

Enhanceddata,surveyquality,andmissionversatilityarecost-benefitadvantagestousingautonomousvehicles.

Theprimeadvantageofthesevehicleshoweverwouldbetoreducethedeploymentofconventionalplatforms

and personnel in the field. Hypothetically, environmental monitoring could be completed by autonomous

systemswithout the need for a conventional vessel leaving harbour or an aircraft taking off. The offshore

environmentisinherentlyhazardousandtheimplicationthatanautonomousvehiclecouldremovetheneed

forapersontoleavethesafetyoflandissignificant.Majorconcernsareraisedwithmanaginganyvessel/aircraft

offshoreandconsiderableriskanalysiswouldberequiredforunmannedcraftinanareaofindustryoperations.

Thispotentialreductioninhealthandsafetyrisksislikelytobeextremelyappealingtoindustry.

However,thepossible introductionofalternativehealthandsafetyrisk intoat-seafieldofoperationswould

delaythematurationofautonomousvehiclesinanE&Poperationalsetting.Ahighlevelofunderstandingand

confidencewillberequiredregardingoperationalhealthandsafetyrisks.Proceduresformanagingunmanned

deviceswouldneedsignificantdevelopmentanddetailedriskassessmentsandwouldneedtobeconductedon

acasebycasebasis.

20 https://bts.fer.hr/_download/repository/Jose_Pinto.pdf

118


DATE:July2016


Safetywillbetheoverarchingthemeasweconsideroperationalpracticalitiesofutilisingautonomousvehicles

in an industrial setting. The vast range of possible contexts requires that we take a general approach in

attemptingtocompareacrossplatforms.Thecategorieswithinwhichwewillexploretheseissuesare:

• TechnicalReliability:Extensivetrialsandproventrackrecordarelikelytoberequiredofanycraftbefore

usageinindustrialfieldofoperations.

• MissionDuration: Any autonomous vehicle that canmaintain longmission durationwith complete

independencewillofferacrucialadvantage.

• Logistics:Logisticalfactorsmustbetakenintoaccount,albeitonaprojectbyprojectbasis,forsafety

concernsandalsocostimplications.

• RemoteOperation:Technicalandsafetyassurancewillberequiredregardingthedependabilityofthe

linkandresponsiveabilityofremoteoperation.

• Interaction with other marine/airspace users: Clear code of practice is likely to be required and

disseminatedtopotentialinteractorsintheareainadvance.

Additionally,practicalrealitiesofoperations,suchasdeployingandretrievingthecraft,tendtofalloutsidethe

scope of studies and very little published informationwas found exploring these issues. However, practical

knowledgegainedfromfirst-handexperienceoftrialsisincluded.

9.3.1 TechnicalReliability

Confidence in reliabilitywouldbe required inmechanicalandmanoeuvrabilityperformance.Any riskof, for

instance,ASVbreakdownaheadofaseismicvesselandassociatedequipment,wouldhavetime,costandsafety

implications.Supportmightbeavailableviaachasevesselbutthisisnottobeguaranteed.Closeattentionto

accuratewaypoint-markingandremoteoperation,includinganover-ridefacilityincaseofanymalfunctionis

required.Ofthehighestimportanceisthatanyautonomousvehicleoperatinginthevicinityhassufficientpower

andmanoeuvrabilitytobewhereitneedstobeattheappropriatetime(oratleastnottobewhereitshouldn’t

be).Absolutelykeytoindustryoperationsisthatanyadditionalhardwaremustnotinterferewiththeindustrial

operations taking place. In the case of ASV, trials to date have invariably involved close support from an

accompanying(conventional)vessel.ThestageissetforanASVmodeltoproveitscapabilitytoindependently

completeextendeddurationmissions.Forunderwaterbuoyancygliders,Brito,(2014)calculatedriskprofilesfor

shallow and deep gliders (combiningmanufacturer’s platforms). They found that the probability of a glider

survivingitsmissiondeploymentwithoutaprematuremissionendwas0.5and0.59fordeepandshallowgliders

respectively.Thisidentifiesthatcurrentlyotherfactorsthanbatterypowerarelimitinggliderperformance.Such

analysesshouldbecompletedforallautonomousplatforms(AUV/ASV/UAS)intendedforsurveysconducted

fortheO&Gindustry,astheyidentifytheamountofredundancyrequired(intermsofnumberofvehiclesto

completethemission),aswellasinformonlikelihoodoftechnicalfailure.Onecurrentchallengeisthatmanyof

119


DATE:July2016


theseplatformsareconstantlyevolving.Extensivetrialsandproventrackrecordarelikelytoberequiredofany

craftbeforeusageinindustrialfieldofoperations.

9.3.2 MissionDuration–Powerconsiderations

Themeansofpowerunderpinsmuchoftheoperationalaspectsdiscussedinthisreview.Therearearangeof

power methods available (often as hybrid options on the same craft). Additionally, the power budget is

frequentlytailoredtoprojectspecifics.However,inbroadterms:conventionalfuelrequiresrefuelling,which

limitsmissionduration.Renewableharvesting(e.g.solarenergy)addressesthisbutrequiresaconsistencyin

supplythatcannotbeguaranteedduetovariabilityinenvironmentalconditions.Powerbywavemotionoffers

thegreatadvantageofalmostlimitlessmissionduration,albeitatlowspeedsandlowermanoeuvrability.The

meansofrefuelling/rechargingalsorequiresconsideration.InthecaseofASV/AUV,asupportvesselwould

berequired,incurringmarineoperationlogistics,costsandrisksthatanyASV/AUVintendstoreduce.Inmany

instances,particularlyforUAS,themissiondurationwilldefinetheoperation.Withinthisdefinedmissiontime

limittherewouldhavetobesignificantcontingency.Anautonomousvehiclethatcangenuinelymaintainlong

missionduration,whollyindependentlyofanysupportvessels,offersacrucialadvantagebyremovingtheneed

forin-fieldlogisticalsupport.

9.3.3 Logistics

Thedeploymentandrecoveryofautonomoussystemsisanimportantconsideration.Completeindependence

wouldbepreferred,wherebyacraftexitsandarrivesatasurveysiteatseaunderitsownpower-apossibility

forsomeASVandAUV,andmanyunderwaterbuoyancygliders(albeitslowlyinthelattercases).Shortofthis

capability,mechanismssuchasasupportingvesselfittedwithliftingequipmentandsuitablelandingareaswill

berequired.ThesewillbringfurtherHSEconcerns,especiallyinpoorweather.Transportationofcraftisalsoa

considerationtobemade.Thereisahugerangeofsizesinautonomousvehicles(Table13)butallwouldlikely

berequiredtobefreightedtotheoperationsarea.Thismightwellbeoverseasandissuesincludingimportation

requirementsandinparticulartransportationoflithiumbatteriesmaycomeintoplay.Logisticalfactorsmustbe

takenintoaccount,albeitonaprojectbyprojectbasis,forsafetyconcernsandalsoimplicationsoncost.

9.3.4 RemoteOperation

A large range of, often sophisticated, remote control mechanisms and options are covered through the

unmannedvehiclesreviewedhere.Howtheywouldbeimplementedinpracticeinanoperationalareaneedsto

beaddressed.Theoperationofavehiclefromaremotelocation,suchasbysatellitelink,opensthewayforcraft

beingmanoeuvred in an E&P survey area with several other vessels or installations on-site by a specialist

thousandsofmilesaway.Communicationprotocolswillbevital,toensurethat“ground-truthed”information,

especiallynavigationalandenvironmental, issuppliedtoassisttheremoteoperatorandfull transparencyof

120


DATE:July2016


suchremoteoperations,not leastpositioninginformation, istransmittedtotherelevantpersonneloffshore.

Thisissueadditionallyimpactsinteractionwithothermarine/airspaceusers.

Thecontrol,handlingandmaintenanceofautonomousvehiclesmayrequirespecialistknowledge.Thiswould

therefore entail the provision of specialists into the offshore working environment. A key implication of

additionalpersonnelisphysicalbunkspace- invariablyatapremiuminoilandgasoperations.Alternatively,

specialisttrainingcouldberolledoutforexistingcrew.

Technicalandsafetyassurancewillbe required regarding thedependabilityof thecommunication linksand

responsiveabilityofremoteoperation.

9.3.5 Interactionwithothermarine/airspaceusers

Safety isparamountwith thenumberanddiversityofotherassets ina fieldofoperation. Theoperationof

autonomousvehicleswithinanindustrialfieldofoperationsmeansthatit’shighlylikelythatotherplatforms

willbeinthearea.Also,thehardware(whetherdeployedfromexistingvesselsorfromautonomousplatforms)

must not interferewith the industrial operations taking place. Of the highest importance is, asmentioned

previously, that any autonomous vehicle is technically reliable in itsmanoeuvrability. Smaller, lowpowered

vehicles,whichareeasiertodeploy,arelikelytohaveproblemsinthisarea,inwhichcasemeasureswouldneed

tobetakentorapidlyrecoveranyvehiclewhich isposingahazardtootheractivities.Thoughgovernmental

regulationsforsafeusemaybeinplace,theseareinearlystageofdevelopment(especiallyincaseofmaritime

activities).Considerationwouldneedtobegiventothespatialrequirementsofindividualautonomousvehicles

inanygivensetting,relativetoothermarine/airspaceusers.Sensorsandmechanismsforcollisionavoidance

willbeofhighpriority,aswillmethodsforpositioningandreportingthatposition.Datatransmissionsystems

will have to comply with local regulations and restrictions and there should be minimised risk of cross-

interference.Remoteoperationisakeycomponentherewiththequestionarisingofwhetheraremoteoperator

isrequiredtocommunicatewith,potentially,thirdpartymarine/airspaceusers.Ifso,themeansbywhichto

achieve such communicationwith thirdpartyuserswouldneed tobedefined. In summary, a clear codeof

practiceislikelytoberequiredanddisseminatedtopotentialinteractorsintheareainadvance.

9.4 Regulatory/politicalbarriers

9.4.1 Regulationsfortheairspace

ThereareseveralregulatoryandpoliticalconsiderationsinvolvedintheuseofUASinoffshoreoperations.For

instance,studysitesnearairportsorpopulatedareasandnocturnalsurveysmaybemorerestrictedthanremote

sitesanddaylightsurveys,andmaydemandspecialpermitsbylocalauthorities.However,regulationsgoverning

the use of UAS are currently changing, and users should seek updated information in due time before the

plannedresearchoperations(Wattsetal.,2010).Withoutanon-boardpilot,thereisasignificantrelianceon

the commandand control link, anda greater emphasison the lossof functionality associatedwith lost link

121


DATE:July2016


communicationwiththegroundstation.Furthermore,forAirTrafficControl(ATC)operationsrequiringvisual

meansofmaintainingin-flightseparation,thelackofanon-boardpilotdoesnotpermitATCtoissueallofthe

standardclearancesorinstructionsavailableunderthecurrenteditionofFAAOrder7110.65,ATC(FAA,2013).

Consequently,toensureanequivalentlevelofsafety,UASflightoperationsrequireanalternativemethodof

compliance(AMOC)orriskcontroltoaddresstheir“see-and-avoid” impedimentstosafetyofflight,andany

problemstheymaygenerateforATC.Currently,thelackoftechnologytodetectandavoidothervehiclesintheir

vicinity,isoneofthemainlimitationsfortheuseofUAS,which,ifimplemented,wouldbehighlyvaluablein

preventingcollisionswithotheraircraftparticularlyduringoperationsbeyond-Line-Of-Sight(BLOS).Infuture,

permanentandconsistentmethodsofcompliancewillbeneededforUASoperationsintheNationalAirspace

(NAS)withouttheneedforwaiversorexemptions(FAA,2013).

Depending on the type of operation and the air space regulationswithin the country of operation, certain

requirementsmustbefulfilled(seeFigure13).InNorway,forinstance,allapprovedRemotelyPilotedAircraft

System(RPAS)operatorscanoperatewithinVLOSandtypicallybelow400ftwithoutapplyingforapermitfrom

the civil aviation authority. For a VLOS operation, visual contact without any optical devices, other than

prescription glasses,must bemaintained at all time. Themaximum VLOS (vertical) distance from the pilot

dependsonweatherconditions,timeofdayandtypeofbackground.Itcanvaryfrom200to300mforsmall

multi-rotorsandupto3to4kmforfixedwingaircraftsfittedwithhighvisibilitypaintorlighting21.Itispossible

toextendthisrangebyutilisingoneormoreremoteobserverssothattheunmannedaircraftiswithinvisual

sightofanobserveratalltimes.Theoperationisthenreferredtoasanextendedvisiblelineofsight(EVLOS)

operation. Critical flight information is relayed via radio for assisting the remote pilot in maintaining safe

separationfromotheraircrafts.RPASoperationsbeyondvisiblelineofsight(BLOS)isusuallyonlyapprovedin

areaswheresafeseparationtootheraircraftscanbeensured,suchasremoteorsegregatedairspace.BLOS

operations require an approval from the civil aviation authorities, which has to be applied for in advance

(typically3to4monthsinNorway).

Currently,allUASoperatorsshouldensuretheiroperationsmeetallapplicableNationalCivilAviationAuthority

(NCAA),local,stateandfederalregulatorylegalrequirements.TheIOGPlaunchedasetofguidelinesinearly

2015, specifically designed forUAS operations. These guidelines include issueswhich are highly relevant to

ensuresafeaerialoperationssuchasmaintenance,communication,andtraining(FAA,2013).Theseguidelines

alsospecifythatoperationsshouldbeincorporatedintoaSafetyManagementSystem(SMS)consistentwith

theOilandGasProducer’sAircraftManagementGuidelines(AMG)(Luftfartstilsynet,2015).

ThemainlimitationtotheuseofUASistheacceptanceofthetechnologybyregulatorybodies,aviation-related

restrictions on flyingUAS and responsiveness by systemmanufacturers. It is yet to be assured that current

monitoringandmitigationmethodscanbereplacedbyUAS,andthisisoneofthecriticalpointsthatstimulate

21 http://www.acuo.org.au/industry-information/terminology/how-do-we-see-them/

122


DATE:July2016


moreresearchanddatacollectionaboutdifferentsystems.Atthesametime,theneedforfurtherresearchis

limitedbytheavailabilityofpermitsandregulationsthatallowsit.However,countrieslikeNorway,Canadaand

Australiahaveinvestedindevelopingnewsystemsandnewmethodsofsurveillance,whichinturncanfacilitate

theacceptancebymorerestrictivejurisdictions.ThishasalreadystartedtochangeincountriessuchasUSA,

wherethemainaviationauthority,theFederalAviationAdministration(FAA),hasstartedtotakemeasuresfor

safeUASintegrationintheirregulations.

Country-specific regulations are under development, considering mainly commercial applications. The

continuousfocusonairspaceregulationsthataremoreintegrativeofUASworldwideisaclearsignthatthis

technologyisheretostay.

Figure13.CurrentcommercialsmallUASrulesincountrieswithstrongandgrowingmarketsforUASsystems(Bonggay,

2016)22.

9.4.2 Regulationsforthesea

Theregulatoryframeworkformarineautonomousvehiclesisinastateofflux.Thereisaclearneedforgreater

clarity around legal issues and diplomatic clearance as applied for the different users, e.g.military,marine

scientificresearchor-increasingly-commercialsurveys.Unmannedsystemsarebecomingubiquitousinthe

oceans.Internationallawgoverningactivitieson,overandundertheseaemergedwellbeforethedevelopment

22 http://media.precisionhawk.com/topic/commercial-drones-faa/ (accessed 2/10/2015)

123


DATE:July2016


of unmanned systems. As ASV and AUV become more advanced – already autonomous, expendable and

“intelligent” robotsareemerging, formingnetworkedsystems– legalandpolicy issuesarebecomingacute.

Legitimate concerns over insurance and recoverable loss are also raised. Kraska (2010) points out that as

“aircraft” and “vessels,” unmanned naval systems fit within the existing legal architecture for peacetime

maritimeoperations, including the 1944 Chicago Convention on Civil Aviation and the 1982UnitedNations

ConventionontheLawoftheSea.Thesetwotreatiesandtheirprogenyprovideguidancefortheuseofmostof

the global commons, and reflect a liberal legal architecture for unmanned air/sea vehicles and systems.

Furthermore,the1982UnitedNationsConventionontheLawoftheSea(UNCLOS)playsakeyroleinthelegal

frameworkforAUV/ASVdeployment.However,theavailableStatepracticeconcerningthedeploymentand

operation is not uniform or universal in scope with regard to customary international law

(CORDIS_result_171941).EventhelegaldefinitionofanAUV/ASVremainsunclear,i.e.whetherornotacraft

isclassedasa“ship”(GROOM,2013;Wynnetal.,2013).

Asaresult,thelegalviewsrelatingtotheoperationaluseofautonomousmarinevehiclesarevariedandtend

to be driven by scenario (e.g. requirement for navigational safety) or by class of operator

(defence/academic/commercial)(Rogers,2012).Assuch,itisimportanttotrackcurrentprogressinforminga

pragmaticCodeofPracticefortheuseofAUV/ASV,alsoknownasMaritimeAutonomousSystems(MAS).At

thepresenttime,suchcraftaretypicallyrelativelysmallvessels,operatingonboththesurfaceandunderwater

(ASVandAUV).Wenotethatresearchiswellunderwayonmuchlargercommercialvessels,ledbycompanies

suchasRollsRoyceandtheEUMUNINproject.

IntheUK,theMASRegulatoryWorkingGroup(MASRWG)isaddressingthisspecificneed.Itwasformedin2014

withtwomainaims:

4 ToformulatearegulatoryframeworkthatcouldbeadoptedbytheUKandotherStates,aswellasthe

internationalbodieschargedwiththeresponsibilitytoregulatethemarineandmaritimeworld;and

4 TodevelopaCodeofPracticeforthesafeoperationofMAS.

AninitialInformationPaper,preparedbytheMASRWG,waspresentedbytheUKtotheInternationalMaritime

Organisation(IMO)inJune2015.Thiswasacceptedasasignpostforothernationstobecomeawareofthework

the UK is undertaking on the combined requirements for regulation, equivalence, training, standards and

accreditationforASV.FollowingasuccessfulconferenceinOctober2015,thenextoutputoftheMASRWGwill

betopresentafullCodeofPracticetoIMO,throughtheUKMaritimeandCoastguardAgency.Thiswillbeat

theMaritimeSafetyCommittee(MSC)96inJune2016.

AnumberofothernationsandorganisationsarenowliaisingwiththeMASRWGto inputtotheUKCodeof

Practice;internationalconsensusisakeycomponentofthiswork.Inpractise,anincreasingnumberofAUVand

ASVarebeingoperatedindifferentpartsoftheworlddespitethelackofformalcountryspecificregulations.

Craftoperate,essentiallysafely,byworkingwithintheexistingconventionsandregulationsasfaraspossible.

124


DATE:July2016


TheseincludetheCollisionRegulations,UnitedNationsConventionontheLawoftheSea(UNCLOS),Safetyof

LifeatSea(SOLAS),InternationalConventionforthePreventionofPollutionfromShips,1973asmodifiedbythe

Protocolof1978(MARPOL),NoticetoMariners,NavigationalTelex(NAVTEX)andmanymore.Riskassessments

and safety cases are a critical part of their safe operation. Understandably, ASV and AUV must earn the

confidenceofthewidermaritimecommunitytobeacceptedassafemarine-users.Toprovidesuchassurance,

hugedevelopmentalefforthasbeeninvestedintomethodsfornavigationalpositioningandcollisionavoidance.

Technologysuchasautomaticobstacleavoidance isregardedaskeytopavingthewayforestablishing legal

policiesforunmannedvessels(Campbelletal.,2012).Variousgapsintheexistingdocumentationbringscope

forconfusion,particularlywithintheplethoraofdefinitionsthatareemerging.Agoodexampleofthisiswith

thevariousdefinitionsandinterpretationsoftheword‘Autonomy’itself.TheCodeofPracticewilladdressthis

andmanyotherissues.ItmayprovelargelyunnecessaryfortheemergenceofASV/AUVtospawnasuiteofnew

regulations. Wherever feasible, they would find their place within the existing structure. The principle of

equivalenceiskeyherewherebymarinersofeverykindofvessel,mannedorunmanned,havethefreedomto

operateasdesired-underacommonunderstandingoflegalityandsafety.

TheintentionisthatanIndustry-ledCodeofPracticeforAUVandASVwillhelptomakethistransitionalprocess

assmoothaspossible.Inturn,thiswillfollowonaworldwidescalewithcountry-specificrequirements.

9.5 Technicalchallengesofmanaging,storingandanalysinglargeamountofdata

Overthecourseofaday,autonomoussensorscanacquiretensofGigabytesofrawdata.Themanagementand

sciencequestionstoanswerwiththesedataoftenrequirethereductionthedatatooneortwonumbersor

simpledecisions,e.g.“Isthereananimalintheexclusionzone?”or“Wasananimaldetectedinthe20minute

period?”,“Canwestartthesurveyline?”,“Thedensityofanimalsinthisareais…”.Thejobofthedataprocessing

chainistoextracttheserelativelysimpleanswersfromthegreatvolumeofrawdata.Dataprocessingisoften

doneinmultiplestagesandsensingsystemsoftencombineautomaticprocessingofrawdatawithdatadisplays

viewedbyahumanoperatorasanaidtofinaldecisionmaking.

Sensorpackagesforautonomousplatformscomeinmanyformats.Thesimplestsystemsconsistoflittlemore

thanthesensoritselfandsomesortofdatarecorder(Figure14A).Thesedays,therecorderisoftenalowpower

computerthatstreamsdatafromthesensortoadatastoragedrive.Whilesimple,suchsystemscanhavethe

advantageofhighreliabilityandlowpowerandareoftenidealifdataarenotneededimmediately.Slightlymore

complexsystems,oftenusedwithhighbandwidthdatawherestorageisaproblem,mightcarryoutsomeinitial

screeningof thedata inorder to removethosedatahighlyunlikely tocontainanythingof interestbutkeep

processingtoabareminimumanddonotprovidedatainreal-time(Figure14B).Ifdataarerequiredinrealtime,

thenthosedatamustbetransmittedtoshoreoranearbymannedplatform(vesseloraircraft)sothattheycan

beusedinnearreal-timedecisionmaking(Figure14C).Thiscancreatesignificantchallenges,particularlywhen

workingwithhighfrequency(andthereforehighvolume)dataoverlowpowerorwhereitisexpensivetosend

125


DATE:July2016


largevolumesofdata(e.g.satellite).Whenit isnotpossibletotransmitrawdata,thensufficientprocessing

musttakeplaceonboardtheplatforminordertoreducetheamountofdatatothatwhichcanpracticallybe

transmitted(Figure14D).

Thequantityofdatathatcanbesentthroughthevariouscommunicationsdevicesvariesenormously(section

8.1.4) but suffice to say that many communications options do not have the capability to transmit high

bandwidthdata,inwhichcaseitisnecessarytoprocessdataontheplatformitselfinordertoreducetheamount

ofdatathatistransmitted(Figure14D).

Figure14.Schematicdiagramsoffourdifferentsensordatahandlingcombinations.Theseareintendedprimarilyas

examplesandmanymorecombinationsexist.A)Simplesensinganddatastorage,B)Sensor,basicprocessinganddata

storage,C)Real-timetransmissionofrawdatawithprocessingonshore,D)Onboardprocessingtoreducedatavolume

priortotransmissionofaminimalamountofdata.

Hereinliethefundamentalproblemsandlimitationsofsensorsonautonomousplatforms:forsituationswhere

dataarenotrequiredinnearrealtime,thelimitationsarestoragecapacityandpower;ifdataarerequiredin

real-time(ornearreal-time)thentheamountofdatathatcanbetransmittedaregovernedbytransmission

distance,poweravailabilityandcost.Datamustthereforebereducedbywhateverfactorisrequiredtosatisfy

thoselimitations.However,processingalgorithmsmustbesufficientlyreliableandsufficientdatamuststillbe

transmitted that an operator has enough information for accurate decisionmaking. If insufficient data are

126


DATE:July2016


presentedtoanoperatorandthereisanoverrelianceonpotentiallyunreliableautomaticprocessingalgorithms,

thenerrorswilloccur:animalswhicharepresentmaybemissed,andfalsealarmswillbereceived.

Theamountofprocessingthatcanbecarriedoutontheplatformwilldependonthesensortype,thestrength

anddistinctivenessofthesignalsofinterestandalsothetypeandquantityofbackgroundnoisewhichmayvary

widelybetweenregions.Forsomesensors,speciesandsituations,researchershavebeenabletoreducedata

sufficientlyforreliableinformationtobesenttoshoreviarelativelylowbandwidthsatellitecommunications

links(e.g.Baumgartneretal.,2013).Thisdoesnot,however,meanthatsuchsystemscanbeeasilydeveloped

forotherspeciesandothersituations.Inmanycases,ithasprovennecessarytousemuchhigherpower,short

rangecommunicationssystemswhichtransmithighvolumesofdataoverrelativelyshortdistances.

9.5.1 Electro-opticalimagingsensors

Opticalsensors(RGB,IR,andvideo)maydifferintermsofdatatransferandprocessingdependingonthetype

ofsensorinconsideration.Still-photoandvideodata(eitherinRGBorIR)maybestoredon-boardanSDcard

forpostprocessingorgatheredduring real-time transmissionofavideoandmanually triggered toobtaina

photoofatarget.Theresolutionrequiredheremayplayakeyroleinthetypeofsensortobeused,sinceitis

possibletotransmitHDimagesinreal-time,buthigherresolutionsstillappeartobeunreliableforthistypeof

application.Datatransmissionwillthereforedependonthedatatype,withradiolinksaffectingtherangeto

whichitispossibletofly.Forhigherranges,itisthenrequiredtoacquireastrongerradiolink,suchasKongsberg

MaritimeBroadbandRadio.

Automateddetectionofobjectsofinterestisanadvantageincircumstanceswheretheamountofdatatobe

acquiredisverylarge.Thisishighlyrelevantparticularlyfortransect-basedsurveysinpopulationstudiesand

monitoringof areas relevant to seismicoperations. Zitterbartet al. (2013)have shown thepotentialof this

technology indetecting largecetaceansusing thermal IR systems in shipsurveysby identifyingwhaleblows

basedon their thermal signature.Additionally,workdevelopedbetween LGLAlaska, Inc, andBrainlike, Inc.

(Ireland et al., 2015) has progressed the availability of a real-time animal detection software. This type of

softwarerequiresadegreeofmachinelearning,sincesurveyconditionsmayvarybothintime,space,andinthe

type of species possible to encounter. On the other hand, algorithms that are operational for post-survey

analysis are still unreliable due to the lack of animalmovement that improves detectability.Mejias, (2013)

developedtwoalgorithmsthatseemtobepromising inmarinemammaldetection,however,theamountof

falsedetectionsremainedhigh,highlightingthestatusofdataprocessingforpost-surveyanalysis.

ThoughUASareanupcomingtechnologywithmanydifferentapplications,theamountofdigitalimagerythat

thesesystemscollectcansometimesbeoverwhelming.Dependingonthesurveydesign,thetimerequiredto

manuallyanalyse the images fromeach survey canmake thismethodology costly,highlighting theneed for

automatedanalysisprocesses.Therearemany issuesconcerning theuseofautomateddetectionofmarine

organisms; sea state condition isoneof the key issues affectingdetectability,which in termsof automated

127


DATE:July2016


detectioncouldhaveamajoreffect, since thepresenceof surfwakecancreate falsepositives thatmaygo

unnoticedunlessmanuallychecked.Anotherissueisthepresenceofglareandotherenvironmentalfactorsthat

mayaffectthedetectionofanimalsbasedpurelyonbodyshape.Irelandetal.(2015)developedanautomated

process that can quickly process high resolution digital still images and identify potential marine mammal

detections, with an overall successful detection rate of approximately 70%. The time required tomanually

reviewtheautomatedprocessingoutputbasedonthisworkwaslessthan2%ofthefullmanualanalysistime,

thus significantly reducing the amount of time and associated cost of image analysis. This work, done in

collaborationwithBrainlike Inc.,hasbeen incorporated intoaerial surveyworkconductedbyShellOilusing

PixMin™, a software designed for target recognition. However, most of the work developed in automated

detectioninvolvespostsurveyanalysis,althoughKoskietal.(2013)showedthatreal-timemonitoringwithUAS

andvideocamera is feasible.Theneed for survey technology (platformsandsensors)hasbeenpushing the

developmentofaerial systemsthatcandeliveraccurateandunbiaseddataboth to industryauthoritiesand

research.Therefore,imageryanalysisremainsaconstraintthatisquicklybeingovercomebytheprogressinthe

developmentofbothsensorsandaerialplatforms,asthequalityofthedataimprovesandtheavailabilityof

detectionalgorithmsthatcanbetestedincreases.

9.5.2 PAMsystems

Table9showsthevolumeofrawuncompresseddatafromasinglehydrophonedigitizedwith16bitaccuracy.

Notethatthesearerawdatavolumesandthatitispossibletoruncompressiononaudiodatawhichwillreduce

datavolumebyaroundafactoroffourusingalgorithmssuchasFLACorX3(Johnson,2013).Ifmulti-hydrophone

systemsareconsidered(aswouldberequiredforanimallocalisation)thenthesevaluesshouldbescaledupby

thenumberofhydrophones.As canbe seen, forbaleenwhales,whichvocaliseat low frequencies,ayear’s

recordingwillgenerateonlyamodestamountofdataandevensamplingatthestandardhumanaudiosample

rateof48kHzforayearwillgeneratenomoredatathancanbefittedonasingleportableharddiskdrive.

However,forthemajorityofdolphin,beakedwhaleandporpoiseecholocationclicks,higherbandwidthsare

required,whichgeneratesignificantvolumesofdata.

Eveninsituationswherelargedatavolumescanbestored,transmittingthosedatatoaremoteplatformmay

beimpracticalduetobandwidthandpowerlimitations.Ifsufficientpowerisavailable,radiolinkscanbeused

overrangesofafewkilometres;forlargerdistancesinremotelocations,itmaybenecessarytousesatellite

communicationssystems,inwhichcasepower,bandwidthandcostallbecomeproblematic.Inthiscasetheonly

optionistopre-processdataonboardtheautonomousplatformandtransmitonlyasummaryofthedetection

data,whichishopefullysufficientforaccuratedecision-making.Insomecases(withlowfrequencydataorhighly

distinctivespecies)thishasbeenprovenpossible,howeverforspecieswhicharedifficulttodistinguishfrom

backgroundnoise,orforwhichcomplextrackingisrequired,itiscurrentlynotpossibletotransmitsufficient

datathroughlowbandwidthsystemsforaccuratereal-timedecisionmaking.

128


DATE:July2016


Table9.Datavolume(Gigabytes)fordifferentrecordingbandwidthsandrecordingdurationsforasinglehydrophone

recordedat16-bitaccuracy.

SampleRate(Hz)

Bandwidth(Hz)

GigabytesperDay

GigabytesperWeek

GigabytesperMonth

GigabytesperYear

SpeciesAccessible

250 125 0.04 0.28 1.21 14.69Blue and some other largebaleenwhales

2000 1000 0.32 2.25 9.66 117.48 Baleenwhales

48000 24000 7.72 54.07 231.74 2,819.54Sperm whales, most dolphinwhistlesandsomeclicks

500000 250000 80.47 563.26 2,413.99 29,370.19Dolphin, beaked whale andporpoiseclicks.

Anincreasingnumberofalgorithmsarenowavailablefortheautomaticdetectionofmarinemammalsounds

withconsiderablemomentumtothefieldbeinggivenbybiennialworkshopsfundedbytheUSOfficeforNaval

Researchandothersandpublishedas special issuesof various journals23.However,manyof thealgorithms

presentedintheliteraturehavebeentunedtoveryspecificdatasetscoveringalimitedrangeofspecies,and

nearlyallrequirehumanvalidationoftheiroutput,particularlywhenworkinginavariablenoiseenvironment.

Researchgroupsprocessinglargedatasetsgenerallytrytofindabalancebetweenautomaticprocessingand

humanvalidation,puttingconsiderableeffort into thedevelopmentofdatavisualisation toolswhichenable

operatorstohomeinoncandidatedetectionsorareasofinterestinthedata(e.g.Tritonsoftwaredevelopedby

SCRIPPSWhaleAcousticsLab,theRavenandXBatpackagesfromCornellandPAMGuardfromSMRU).Forsome

soundtypes,detectoroutputdata,containingonlyveryshortsnipsofrealdata,aresufficientfordatachecking.

Forothers,thecontextualinformationthatisonlyavailablefromaspectrogramofalongersectionofrawdata

can bemore important. For example, Figure 15 shows false and realwhistle detections from an automatic

detector bothwithout andwith the full spectrogram information. From thedetectedwhistle contours, it is

impossibletosaywhetherthedetectionsarerealorfalse.Byviewingspectrogramsoftherawdataandalsoby

listeningtothem,itwasclearthatthefirstsoundismechanical(inthiscasethehydraulicsystemloweringthe

outboardengineonavesseldeployingaPAMequippedWaveglider)andthatthesecondexampleisagenuine

animalsound.However,thefull5secondsofrawaudiodatashowninthisexamplewouldrequirearoundhalf

amegabyteofdatarelay,whereasthedetectiondataaloneonlyrequiresatinyfractionofthis.

23For example: Canadian Acoustics, Vol 32, No 2 (2004), Canadian Acoustics, Vol 36. No. 1 (2008), Applied Acoustics, Volume 71, Issue 11, Pages 991-1112 (November 2010), J. Acoust. Soc. Am. Special issue on Methods for Marine Mammal Passive Acoustics , J. Acoust. Soc. Am. 134 (2013)

129


DATE:July2016


9.5.3 AAMsensors

Theamountofdatacollectedbyactiveacousticmethods(singlebeam,splitbeam,omni-directional,wideband

andmultibeamechosounders)isdependentontheresolution(e.g.pingrate,numberofbeams),thecomplexity

ofthedata(e.g.single/multifrequencyorwideband)andtherangecollectedto.Datavolumeiscontrolledby

varyinganyof these three criteria. Forexample,10pingsof120kHz single frequencydata (1.024mspulse

length)toarangeof100mwithanEK80generatesa~300kBdatafile,whereas10pingsofwidebanddata(95

–160kHz,centralfrequencyof120kHz,2.048mspulselength)toarangeof100mgeneratesa~5mBdatafile.

Asaresultofthishighdatavolumecreation,theAAMsensorstypicallystoredatalocallyforretrievalandanalysis

oncetheplatformisrecovered,unlesstheyarepartofacabledobservatory24.Thelogicalsolutionistolookfor

methodstoreduceorcompresstherawdataortopre-processdataon-boardtheplatform(althoughthishasa

payload and power cost to the platform) for transmission of summary statistics.We are not aware of any

publishedexamplesofdatacompression,althoughthereareexampleswhereitisalludedtoduringtheuseof

24 e.g. http://www.interactiveoceans.washington.edu/story/NSF_Ocean_Observatories_Initiative

A

B

Figure15.Examplesofrealandfalsewhistledetectionsbothwithoutandwithfullspectrograminformation.A)A

falsedetectioncausedbythehydraulicsofanearbyengine.B)Adolphinwhistle.

130


DATE:July2016


anEK60onaHuginAUV(Patel,2004)andforreal-timetransmissionoftunadetectionusingtheZunibalTunabel

e7buoy25.

Methodsforthedetectionandestimationofanimalbiomass, inparticularfish,fromactiveacoustics iswell-

developed (e.g. Korneliussen and Ona, 2003; Petitgas et al., 2003). However, like the PAM systems, these

methods are typically tuned to specific features and species and frequently require human validation.Data

processingandanalysis is typicallyundertakenusingeithercommerciallyavailableacousticsoftware2627, in-

housebuilt(e.g.MOVIES3D,citedinTrenkeletal.,2009)ortoolboxesdevelopedforMatlab28.Thenextstepis

toimplementtheprocessingofdata(withaspecificgoal)on-boardanAUVtosendsummarystatisticsbackto

thebasestation.Weareunawareofanycurrentactivitiesthatwouldenablethis,but it isanareaofactive

researchinterest.

TritechInternationalLtdhasdevelopedthesoftwareGeminiSeaTecmarineobjecttrackingandtargetdetection

system29inclosecooperationwithSMRULtd(nowSMRUConsulting)asadatareductiontoolforbehavioural

monitoringandresearch.Italsoallowsforearlywarningofthepresenceofmarinemammalsaroundrenewable

tidaldevices(Hastie,2012;Hastie,2013).Thissystemusesclassificationalgorithmsbasedonvariablessuchas

size, shape and swimming characteristics that provide a probabilistic indication of the identity of individual

targetsasvalidmarinemammaltargets.ResearchersatSMRUarecurrentlydevelopingimprovedalgorithmsfor

theautomaticdetectionandtrackingofmarinemammaltargetsusingtheTritechGeminimultibeamsystem(C.

Sparling,pers.Comm).Thissystemisalsocurrentlybeingadaptedforsharkdetectiontoprovideanalternative

tocurrentsharkdefencesystems.

10 Acknowledgements WewouldliketothankIOGPforfundingthiswork.WearegratefulforthehelpandsupportprovidedbyGordon

HastiewithregardstoAAMsensorsandmarinemammaldetectionaswellastoCarolSparlingforherreviewof

aformerversionofthisreport.

Wewouldalsoliketosincerelythankallcompaniesthatprovidedsysteminformationforinclusioninthisreport:

ALSEAMAR, ASL Environmental Sciences Inc., ASV, BioSonics Inc., Brainlike Inc., C-ASTRAL Aerospace Ltd.,

CheloniaLtd.,CornellUniversity,EmbeddedOceanSystems,ImagenexTechnologyCorp.,KongsbergMaritime

AS, Kongsberg Underwater Technology Inc., Liquid Robotics Inc., Marine Micro Technology Corp., MOST

(AutonomousVessels)Ltd.,NortekAS,OceanInstrumentsNZ,RTSYS,SCRIPPS,SeicheLtd.,Simrad,StAndrews

InstrumentationLtd.,TritechInternationalLtd.,UTCAerospaceSystems,Vemco,WildlifeAcousticsInc.

25 http://zunibal.com/en/product/tunabal-e7-buoy/ 26 Echoview, https://www.echoview.com/ 27 LSSS, http://www.marec.no/english/index.htm 28 http://hydroacoustics.net/viewtopic.php?f=36&t=131 29 http://www.tritech.co.uk/news-article/gemini-seatec-marine-object-tracking-and-target-detection

131


DATE:July2016


11 Appendix 11.1 Listofcriteriaandmetricsforplatform,sensoranddatarelay

TheevaluationcriteriadefinedandexplainedinsectionCriteriaandmetricsweregroupedintothefollowing

categories:general information,technicaldetails,costs,surveycapabilities,operationalaspectsandmanning

requirementsforboththeplatformsandsensors.Thesegeneralcategoriesfacilitatedtheorganisationofthe

hugeamountofinformationgatheredforeachsystem.Below,abulletpointlistofthecriteriaasdefinedabove

aswellasalinktothecomparisonmatrixtableisgivenforaneasiernavigation.

11.1.1 Platform

Criteria/Metric Description

GeneralInformation Table12includesgeneralinformation:

General Includesthetypeofplatform,platformnameandthemanufacturerand/orthedesignersaswellastechnologyreadinesslevelasdefinedinTable10.

TechnicalDetails Table13andTable14includethefollowingtechnicaldetails:

MissionDuration MaximummissiondurationDurationLimitations FactorslimitingthemissiondurationSpeed Minimum,maximumandtypicalspeed

VerticalRange Minimumandmaximumverticalrange(heightabovesea level foraerialvehiclesanddepthforunderwatervehicles)

EnvironmentalFactors EnvironmentalfactorslimitingtheplatformperformanceSize&Weight OperationalandpackagedsizeandweightHazardClass HazardclassasdefinedinTable11NoiseLevel Noiselevelemittedbytheoperatingplatform

LimitingFactors Factorspotentially interferingwith theperformanceof theplatform (other thanenvironmental)

Interfaces TypesofinterfacesAdditionalSensors PresenceorabsenceofadditionalsensorsontheplatformCTD DeploymenttypeofCTDsensor:mounted,towedorwinchdeployment

Costs Table15containsthecostrelatedcriteria:

Purchase Purchaseprice

Rental Rentalpriceoftheplatform,requestedasper3monthsforAUV/ASVanddailyrateforUAS

Operational Operationalcostsintermsofpilotingmanpowerandpilotingtelemetrycosts

Maintenance Maintenance costs and the costs of batteries and/or fuel used to power theplatform

SurveyCapabilities Table16entailsthesesurveyspecificcriteria:

TrackSetting Availability and kind of a track setting option such as programmed waypoints,manualpiloting,bothoptionsoranyotheroption

EnvironmentalLimitations Environmentallimitationstotrackkeepingsuchascurrentorweather(wind,rain)

Self-Correct Availabilityoftheinstrumenttoself-correctitspositionwhendivertingfromtrack

132


DATE:July2016



StationKeeping Abilityforstationkeeping(oratleastperformturnstoessentiallystayinthesameplace)

AutonomousDecisions AbilityofautonomousdecisionmakingwithmultiplesurveymodesClockSync Abilityforclocksynchronisation

OperationalAspects Table17coversthefollowingoperationalcriteria:

Fuel Fueltype

Failsafe Availabilityofafailsafemodeandkindofmode:bearing,location,endpoint,keeptrackorstop,includinganyotherimportantinformation

Detect&Avoid Transponder,detectandavoidcapacityInterface GroundstationinterfacesuchasIridium,RForBluetoothAutonomy Levelofautonomy:waypointsorcontinuouscontrolPayloadPower PayloadpowerinJoulesandPayloadcapacitydimensionsandweightDeploy&Recover DeploymentandrecoveryprocedureManning

RequirementsTable18coversthemanningrequirements:

Manning Number of people/pilots required to operate the platform and the trainingrequirementsnecessary.

11.1.2 Sensor



General forthesensors:thesystemclass,suchasAAM,PAMorVideo,sensornameandthemanufactureraswellastechnologyreadinesslevelasdefinedinTable10.

Technicaldetails Table21includesdetailssuchas:

Size&Weight Operationalandpackagedsizeandweightoftheplatforms

HazardClass HazardclassasdefinedinTable11,whichisespeciallyimportantforshippingandtransportissues

InterferingFactors Factors potentially interfering with the performance of the sensor (other thanenvironmental)

Interface TypeofinterfaceCosts Table22coversthecostrelatedcriteria:

Purchase Purchaseprice

Rental Rentalpriceoftheplatform,requestedasper3monthsforAUV/ASVsensorsanddailyrateforUASsensors

Operational OperationalcostsintermsofoperatormanpowerMaintenance MaintenancecostsperyearIntegration IntegrationofsensorintoplatformSurveyCapabilities Table23andTable24containthecriteriadefiningthesensor’ssurveycapabilities:

EnvironmentalLimitations

Environmental factors limiting optimal sensor performance such as current orweather(wind,rain)

TypeofDataTypeofcollecteddatatounderstandifdataarestoredprocessedwithanon-boardprocessingprocedureorasrawdata

SpeciesID Abilitytoforspeciesidentificationfromcollecteddatatounderstandifthesensorhasalreadybeenusedtoidentifymarineanimalspeciesand,ifso,whichkind

Bearing Abilityforestimatingbearingtoanimalfromasinglevehiclewiththedatacollected

133


DATE:July2016



DirectRange Ability for direct range estimation from a single vehiclewith the data collected,neglectinganypossiblebearingambiguityorswimmingdepthofanimal

HorizontalRange Ability for horizontal range estimation from the animal to a track line / vehiclelocationfromasinglevehiclewiththedatacollected

Real-Time Abilityforreal-timedatatransmissionofanimaldetections

PAMAbility of PAM sensors for estimation of received levels of the detected animalvocalisation or ambient noise levels, or simultaneous CTD data collection (PAMsensorsonly)

ClockSync Ability for clock synchronisation to be synchronised with other sensors and itslimitationstotheaccuracy

OperationalAspects Table25includesthefollowingitems:

Autonomy Levelofautonomy:waypointsorcontinuouscontrolDeploy&Recover Deploymentandrecoveryprocedure

PayloadPower Payload power in Watts as well as internal and external payload capacitydimensionsandweight

Manning

RequirementsTable26coversmanningcriteria:

ManningSuchasthenumberofpeoplerequiredtooperatethesensor,thevesselaswellasmanningrequirementsfordeployingandrecoveringthesensors,andthetrainingrequirementsnecessary.

Furtherinformation Table27givesinformation

Further

Informationontherawdatavolumeandthereal-timedatareductioncapabilitiesofthesensor.Furtherinformationisgivenifsensorsarerecommendedforfuturefield trials formarine animalmonitoringwith regards to E&P activities, relevantreferences(notyetmentioned),linksandfurthercomments.

134


DATE:July2016


11.1.3 Datarelay



GeneralThegeneralinformationrequestedinthedatarelaysectionincludingthesystemclasssuchasIridium,thesystemnameandthemanufactureraswellastechnologyreadinesslevelasdefinedinTable10.

Technicaldetails Table29includesdetailssuchas:

EnvironmentalLimitations Environmental limitations to optimal system performance such as current or weather(wind,rain)

Size&Weight OperationalandpackagedsizeandweightofthesystemPower PowerData Databandwidthinbitspersecond(bps)Transmission TransmissionrangeCosts PurchaseandoperationalcostsPowerConsumption PowerconsumptionRestrictions UsagerestrictionssuchasgeographicalorpoliticalrestrictionsDelay Datadelay

Training TrainingneedsRequirements Platformrequirements

Table10TechnologyReadinessLevelsusedintheevaluationmatrices*

TechnologyReadinessLevel Description

TRL1:Basicresearch Principlespostulatedandobservedbutnoexperimentalproofavailable.

TRL2:Technologyformulation Conceptandapplicationhavebeenformulated.

TRL3:Appliedresearch Firstlaboratorytestscompleted;proofofconcept.

TRL4:Smallscaleprototype Builtinalaboratoryenvironment("ugly"prototype).

TRL5:Largescaleprototype Testedinintendedenvironment.

TRL6:Prototypesystem Testedinintendedenvironmentclosetoexpectedperformance

TRL7:Demonstrationsystem Operatinginoperationalenvironmentatpre-commercialscale.

TRL8:Firstofakindcommercialsystem Manufacturingissuessolved.

TRL9:Fullcommercialapplication Technologyavailableforconsumers.

*AdaptedfromtheEUFrameworkProgrammeforResearchandInnovation,Horizon2020definitions.

135


DATE:July2016


Table11.Hazardclasslevelsrelatedtobatteryorfueltypeofanevaluatedsystem.Oilsorothermaterialusedin

manufacturingofsystemsthatmayhavehazardsarenotconsidered.

HazardClass Description

1 Explosives

2 Gases

3 FlammableLiquids

4 FlammableSolids

5 OxidizingSubstances

6 Toxic&InfectiousSubstances

7 RadioactiveMaterial

8 Corrosives

9 MiscellaneousDangerousGoods

11.2 Shortlistofplatformsandsensors

Thissection includesashort listofthecurrentstateofthearttechnologiesofavailableUASandAUV/ASV

platformsandsensorstechnologieslistedinthecomparisonmatrices(appendix11.1)thataremostpertinent

toO&Goperationsformonitoringmarinemammal,seaturtlesandfish.

Thecollectedinformationandmatriceswereusedtoevaluatethesystemswithregardstotheirperformance

andsuitabilityforoperationduringseismicsurveysandotherE&Pactivitiesandtheirpotentialuseforassessing

theeffectsofoperationsonmarinespecies.

11.2.1 Poweredaircrafts

11.2.1.1 UX5[TrimbleNavigationLtd]

ThisUAShasbeenunder focusmainly formappingand surveillance solutions. Ithasbeen tested in various

environmentalconditions,suchaswind,extensiveheat,andsnow.TheUX5hasincorporatedareversedthrust,

improvingaltitudemeasurementthatwillassistinaccurateandpredictablelandings.Thisconditionisidealfor

professionalsworkinginsmallareas.Theaircraftisdeployedusingacatapultsystem,andrecoveredusingbelly

landing.Themaximumflight time recorded for this system is50minutes.TheUX5 isdeliveredwithahigh-

resolutionstillcamera,whichstoresimageryon-boardforpostprocessing.

11.2.1.2 BRAMORC4EYE[C-Astral]

TheBRAMORC4EYEUASlineisdesignedforoperationswherereal-timeornearreal-timevideoobservationand

surveillancecapabilityisthemainfocus.Thissystemhasanenduranceof3hoursandastandarddatalinkof30

km. Thedeploymentmethod is similar to theUX5 though theBRAMORC4EYEhas an automatic parachute

136


DATE:July2016


landing,whichallowsforasmootherrecoveryoftheequipment.Thisreducestheriskofdamageofboththe

aircraftandsensors.ItcomesdeliveredwithHDvideoorHDvideo+LWIRpayloadfrommanufacturer.

11.2.1.3 ScanEagle[Insitu]

This platform has been used inmarinemammal studies in Australia (Hodgson, Kelly, and Peel, 2013), and

demonstrateditsabilitytoconductflightsatseaandofuseformarinemammalmonitoring.Theaircraftisable

tosendlivevideofeedtothegroundstation,whichcanbestoreddirectlybythereceiverforposterioranalysis

orusedfortimelydecision-making.Thissystemdoesnotrequirenetsorrunwayarerequired,reducingspace

constrictions in choosing launchand landing locations. Insitucandeliveravarietyofpayload turrets for the

ScanEagle,wheresomearemilitaryonly.ForseamammalstudiesthevisibleEO900ortheVisible+MVIR“Dual

Imager”isrecommended.

11.2.1.4 PenguinB[Uavfactory]

ThePenguinBisacommerciallyavailablefixed-wingUAScapableofflightsofover20hoursandwithsimilar

launchingproceduresasforthepreviouslymentionedpoweredUAS.Whatdistinguishesthisaircraftfromthe

remainingisitsmodularcompositestructure,removablepayload,andneedforawiderareatolandsuchasa

runway. Thismay result in limitedoperations if vessel landing is required, though the flight capacityof this

platformisabletoovercomethisobstacleduetoitsenduranceandoperationalrange.Itcanbedeliveredwith

arangeofcameraoptions,orwithoutpayloadforintegrationofthirdpartysensors.

11.2.1.5 Fulmar[Thales]

TheFulmarUAS is classifiedasamini-UASand isahighlyversatile system,with capabilities foravarietyof

missions.It is launchedusingacatapult-basedlaunchersystemandlandedusingnetlanding.Thus, itcanbe

operated and recovered at any time with few limitations concerning weather conditions and terrain

characteristics.Thissystemhasamaximumflighttimeof12hours,andcanbedeliveredwitharangeofsensor

options,orwithoutpayloadforintegrationofthirdpartysensors.

11.2.1.6 Jump-20[ArcturusUAV]

Jump-20isafixedwingUAScapableofverticaltake-offandlanding,withnoneedforlaunchsystemorrunway.

However, Jumpaircraftcanalsobeconvertedtocatapult launchwithasimplewingchange. It iscapableof

conductingflightsupto16hoursunderharshweatherconditions.TheJump-20canbedeliveredwiththeTASE

400gimbals(orthesmallerversions)fromCloudCap,orwithoutpayloadsforintegrationofthirdpartysensors.

137


DATE:July2016


11.2.2 Kites

11.2.2.1 SwanX1Kite[FlyingRobotsSA]

TheSwanX1Kite(FlyingRobotsSA)consistsofadeliverytricycleequippedwithasoftwingfollowingtheconcept

oftwo-seatermicrolight.Thissystemiscomposedby:

• AnaerialplatformFRSWANXI;

• Anautopilotthatallowsfortake-off,flight,andlandingautonomously;

• A ground station that allows communication between ground operators via UHF/VHF data in case

manualcontroloftheaircraftisneeded.

ThisUASisstillademonstrationsystemanditisnotyetavailablecommercially.However,furthertestingwill

indicatethesuitabilityofthisequipmenttobeaccessibleforfuturescientificendeavours.

TheSwanX1isaversatilesystemthatcanbeequippedwithdifferentsensorsthatareintegratedinthehardware

and software platform. It is capable of carrying a video kit and omnidirectional antenna (supplied by the

manufacturer),whichallowsforreal-timeimagetransmissiontothegroundstation.

11.2.3 Lighter-than-airaircraftsystems

11.2.3.1 OceanEye[MaritimeRoboticsAS]

TheOceanEyeballoon,producedbyMaritimeRoboticsAS,isahelium-filledballoonabletocarryasensorunit

forpersistentaerialsurveillance.Itisatetheredaerostatsystemthatlocalizesthesurveillancetechnologyon-

boardamissionvessel.Inotherwords, it isavehiclewithanaerostaticorbuoyantliftthatdoesnotrequire

movementthroughthesurroundingairmass.Itisdependentonthepresenceofasupportingvesseltowhichit

is tetheredandtransmits its imageryand flight information.This systemhasnotyetbeentested foranimal

monitoring; thoughmost of the applications of theOceanEye are conducted at sea for petroleum-industry

relatedoperationssuchasoilspillmonitoring.

11.2.3.2 DesertStar10[AllsoppHelikitesLtd]

Helikiteaerostatsareacombinationofaheliumballoonandkite (alsoknownaskitoons),whichareableto

overcometheshortfallsofnormaltetheredballoonsandkites.TheHelikitesaretetheredballoonswithawing

thatallowsittosustainharshwindconditions.Theseplatformshavebeentestedinseveralconditionsandseem

tobeveryversatileintermsoftheiradaptabilitytoenvironmentalconditions,areavailableinvarioussizes,and

arereadytotakeondifferentchallenges.

138


DATE:July2016


11.2.4 Propellerdrivenunderwatercraftandpoweredsurfacecraft

11.2.4.1 REMUSFamily[Hydroid,KonsbergMaritime}

TheREMUSisapropeller-driven,torpedo-shapedAUVdevelopedbyHydroid,aKonsbergMaritimecompany.

ThedifferentREMUSmodelshavebeenextensivelytestedincommercialandnavaloperationsinabroadrange

ofenvironments,fromshallowanddeepwaterstoarcticandtropicalregions.REMUS100isacompact,man-

portableAUV,thesmallestofallmodelsavailable;REMUS600isamid-size,highlyversatileandfullymodular

AUV;andREMUS6000isthelargestmodel,designedfordeepwateroperations.

AllmodelsarebatterypoweredanduseaDCmotorattachedtoa2-3bladepropeller.Theycantravelatan

approximatemaximumspeedof4knots,withmissiondurationsbetween8hours(fortheREMUS100)to24

hours(fortheREMUS600).Thevehiclesareoutfittedwithasetofstandardsensors,whichincludeanacoustic

Doppler current profiler (ADCP), a sidescan sonar (SSS), a CTD device, long baseline transponders (LBL), an

inertialnavigationsystem(INS)anddatacommunicationdevices(acousticmodem,IridiumsatelliteandWiFi).

The standard sensor configuration varies between models; additional sensors like cameras, environmental

samplersoradvancedsonardevicescanbeinstalledinthevehicle.TheREMUSfamilyusesacommonsoftware

application(VIPorVehicleInterfaceProgram)formissionplanning,visualizationofsurveyprogressandsystem

statusmonitoring.

11.2.4.2 HUGINFamily[KonsbergMaritime]

The HUGIN is a propeller-driven AUV developed jointly by KonsbergMaritime and the Norwegian Defence

ResearchEstablishment(FFI).Currently,therearethreemodelsavailablecommercially:HUGIN1000,HUGIN

3000andHUGIN4500.Allofthemsharethesametechnologybaseintermsofnavigation,control,payload,

communication, propulsion and emergency systems; themain differences are size, battery technology and

endurance, and sensor configuration. TheHUGIN has a long history of successful operations in commercial

applications, with more than 150.000 km covered in a period of ten years. It was designed to satisfy the

requirementsofcivilianandnavalapplications, forwhichrobustness,dataqualityandoperationalefficiency

werethemaingoals.

The three models use a large blade propeller for high electrical-to-hydrodynamic efficiency and low noise

emission;thenavigationspeedsrangebetween2and4kts(excepttheHUGIN1000,whichcanreach6kts).The

AUVcanoperateinthreemodes:autonomous,semi-autonomousandsupervised.Thenavigationiscontrolled

byanadvancedreal-timeaided inertialnavigationsystem(AINS),whichutilises information froman inertial

measurementunit(IMU),Dopplervelocitylog(DVL),depthsensor,ultra-shortbaselinetransponders(USBL)and

GPS.Datatransmissionandcontrolfromthemothershipcanbesetthroughanacousticmodem,radioorWLAN

link,orIridiumsatellitecommunications.TheHUGIN1000isequippedwithaLithiumpolymerbatteryof54MJ,

whichprovidesanenduranceof24h,lessthanhalfofwhattheAlHPsemi-fuelcellbatteriesofthelargermodels

139


DATE:July2016


canoffer.Themodelnumbergivesthemaximumoperationaldepthinmeters.Thestandardsensorsinstalledin

thethreeAUVareessentiallythesame,butthelargermodelscanbeoutfittedwithamorecapablesensorsuite.

Thestandardsensorconfigurationincludesamulti-beamechosounder(MBE),asidescansonar(SSS)orsynthetic

aperturesonar (SAS),asub-bottomprofiler (SBP),anacousticDopplercurrentprofiler (ADCP),aCTDunit,a

turbiditysensorandacamera.

11.2.4.3 MUNIN[KongsbergMaritime]

TheMUNINisacompactpropeller-drivenAUVdesignedbyKonsbergMaritimeforhighqualitydataandposition

accuracy.TheMUNINcombinesthemostrelevanttechnologyofHUGINandREMUSfamilies,alongwithsome

newdevelopments.Thetorpedo-shapedvehicleconsistsoffivemodularsections:1)thenose,withconductivity-

temperature(CT)andforward-lookingsonar(FLS);2)replaceablebatterymodule;3)navigationandpayload

module;4)controlandpermanentbattery;5)tailwithcNODEelectronicsandtransponder.Thevehicleusesthe

sameoperatingsoftwareanduserinterfaceasHUGINmodels.

TheMUNINincludestwo18MJbatteries,oneofthemremovable,foranenduranceof12to24hours.TheAUV

reachesamaximumspeedof4.5kts.Theinformationofaninertialmeasurementunit(IMU),aDopplervelocity

log(DVL),GPSandsingleormulti-transponderacousticcommunications(UTP,SSBL(SuperShortBaseline),SBL

(ShortBaseline)andLBL(LongBaseline))isassimilatedbyareal-timeaidedinertialnavigationsystem(AINS)for

accurate navigation. The communication is established through an acoustic link (cNODE),WiFi and Iridium

satellite. The suite of sensors available for the MUNIN include a multibeam echosounder (MBE), a dual-

frequencysidescansonar(SSS),aninterferometricsyntheticaperturesonar(ISAS),asub-bottomprofiler(SBP),

aconductivity-temperatureunit(CT)andastillcamera.

11.2.4.4 BluefinFamily[BluefinRobotics]

The Bluefin is a propeller-driven AUV developed by Bluefin Robotics for civil, commercial and military

applications.FivemodelsformtheBluefinfamily:thecompact,two-manportableAUVBluefin9and9M;the

modularandversatileAUVBluefin12Dand12S;andthehighpayloadcapacityAUVBluefin21.Bluefin9,9M

and12Sarebettersuitedforinshoreoperations,duetotheirlimiteddepthrangeof200to300m;theBluefin

12DandBluefin21canbeoperatedindeepwaters,upto4500mforthelastmodel.

Each model is powered with a different number of Lithium-Polymer batteries, according their weight and

requiredendurance.The9and9Mmodelscanbesentona10to12hourmission;thelargermodelsofferan

improvedenduranceof25-30hours.TheBluefin9and12Sareequippedwithasimplernavigationsystembased

onaninertialmeasurementunit(IMU),aDopplervelocitylog(DVL),asoundvelocitysensor(SVS),acompass

andaGPSunit;therestofthemodelsuseaninertialnavigationsystem(INS)combinedwithIMU,DVLandSVS

units.Allmodelsreachamaximumspeedof5kts.Thecommunicationssystemiscommontoallmodelsand

includesanRFlink,acoustictrackingandIridiumsatellitecommunications.Thesensorpayloadvariesfromone

140


DATE:July2016


model toanother,but typicalavailablesensorsaredual-frequency, interferometricanddynamically focused

sidescansonars(SSS),syntheticaperturesonar(SAS),multibeamechosounder,conductivity-temperatureprobe

(CT),backscattersensorandturbidityprobe,amongothers.ThesoftwareprovidedwiththeAUVprovidesan

interfaceforallphasesofamission:planning,execution,monitoringandpost-processing.

11.2.4.5 AFamily[ECAGroup]

A9-MandA18-Darepropeller-drivenAUVdevelopedbytheECAgroup.AlthoughtheAfamilyofAUVconsists

ofseveralmodels,thetwopresentedinthissectioncoverthetechnologyrequirementsformostapplications.

TheA9-M is a compact,man-portable vehiclewith lowmagnetic and acoustic signatures designed to cover

depth-rangesofupto200m;theA18-Disamid-sizedeepwaterAUVwithanextendedenduranceandflexible

sensorpayload.

With3timesmorepowercapacitythentheA9-M,theA18-Doffersanenduranceof24h,comparabletothe

maximum20hoftheA9-M.Themaximumspeedinbothmodelsapproximates5kts.Thenavigationsystem

includesand inertialmotionsensor (INS),aDopplervelocity log(DVL),GPSand, intheA18-D,anultra-short

baseline unit (USBL). Both models are equipped with RF link, WiFi, acoustic tracking and Iridium satellite

communications.ThepayloadcapacityofA18-DishigherthaninA9-Dandsupportsalargernumberofsensors,

whichareoptionalforthecompactmodel.Thesuiteofpossiblesensors,bothstandardandoptional,includesa

singleordual-frequencysidescansonar(SSS),asoundvelocityprofiler(SVP),aforward-lookingsonar(FLS),a

multibeam echosounder (MBE), a CTD unit, environmental sensors, a sub-bottom profiler (SBP) and an

interferometricsidescansonar.

11.2.4.6 ARTEMI,ACESandAutoCat[MIT]

ARTEMIS,ACESandAutoCatwerethefirstautonomoussurfacevehicles.DesignedanddevelopedattheMIT

SeaGrantCollegeProgrambetween1993and2000,theseearlydesignsinspiredotherASVprogramsbeyond

MIT[Manley,2008].Thegoalwastodevelopalightautonomoussurfacevehicletoperformasanavigationand

communicationlinktoanAUV,capableofaccuratesurveyingandsuitableforeducationalpurposes.

ARTEMIS is a 1.37 m long scale replica of a fishing trawler developed in 1993, originally designed to

autonomouslycollectbathymetrydataintheCharlesRiver, inBoston(Manley,2008;Vanecketal,1996). Its

smallsizelimiteditsenduranceandseakeeping,butmadeiteasytobetransported,deployedandrecovered.

Thevehiclewasequippedwithanelectricmotorandservoactuatedrudderandincludedautomaticheading

control and DGPSway point navigation. The installation of a radiomodem allowed for human supervisory

control.

ACES (AutonomousCoastal ExplorationSystem) is a small catamaran1.4m longand0.4mwidedeveloped

between1996and1997.Thevehiclewasoutfittedwithhydrographicsurveysensors.Moreversatileandstable

thantheARTEMIS,theACESwasequippedwithtwocommercialhullslinkedbyasteelstructure,navigationand

141


DATE:July2016


controlsystems,a3.3hpgasolineengineforpropulsion,batteriestoprovidepowertothecomputers,anda

generatortorechargethebatteries.Thevehiclewascharacterisedbycruisingandmaximumspeedsof2and

2.25kts,a30kgpayloadand4hoursendurance(Manley,1997).

AutocatistheupgradedversionoftheACEScatamaranandwasfirsttimetestedinthesummerof2000(Manley,

2000;2008).AutoCatis1.8mlongandcantravelatspeedsof8m/sthankstoanelectrictrollingmotor,which

replacedtheoriginalgasolinepropulsionsystemoftheACES.PayloadsensorsincludeDGPSandoceanographic

equipment,butcanbeoptionallyoutfittedwithasub-bottomprofiler,communicationslinkandSONAR.Last

trialswerecarriedoutin2000anddespitetheirhistoricalvalue,itisunlikelythatthesemodelsarestillavailable

foroperation.

11.2.4.7 MeasuringDolphin[MESSIN]

TheMeasuringDolphiniscatamaran-shapedASVwithfiberglasshullsdevelopedundertheMESSINprojectin

Rostock(Germany)andsponsoredbytheGermanBMBF(Cacciaetal.,2009;Caccia,2006).Designedfortrack

guidanceandcarryingmeasuringdevicesinshallowwater,thisvehiclewasequippedwithhighlyaccurateDGPS

positioning,compassandautomaticcoursecontrol.TheMeasuringDolphinwasequippedwithonepropeller

on each hull and a hybrid energy supply system – batteries and internal combustion for electric power

generation.

11.2.4.8 Delfim[DSORLab]

TheDelfimisasmallcatamaran-shapedASVdesignedanddevelopedbytheDSORlabofLisbonIST-ISR,under

theEUfundedprojectASIMOV(AdvancedSystemIntegrationforManagingthecoordinatedoperationofrobotic

OceanVehicles) (Alves, 2002; Caccia et al., 2009; Caccia, 2006; Pascoal, 2000) This vehiclewas designed to

performautomaticmarinedataacquisitionandtoserveasanacousticcommunicationrelaybetweenanAUV

and a support vessel. TheDelphimwas equippedwith an 80 km rangeRF link for communicationwith the

supportvessel,afixedGPSstationandtheon-shorecontrolcenter.

11.2.4.9 Springer[UniversityofPlymouth]

TheSpringerisacatamaran-shapedASVdevelopedbytheUniversityofPlymouth(UK)fortracingpollutants.

Thevehicle is 3m longand1.5mhigh. SpringerwasdevelopedbyMarineand IndustrialDynamicAnalysis

ResearchGroup(MIDAS)atPlymouthUniversityasacosteffectiveandenvironmentally friendlyASV. Itwas

designedprimarily forundertakingpollutanttracking,andenvironmentalandhydrographicsurveys inrivers,

reservoirs,inlandwaterwaysandcoastalwaters,particularlyshallowwaters.Springeralsoservesasatestbed

platformforresearchinintelligentnavigationsystemsandsensorandinstrumentationtechnology.

142


DATE:July2016


11.2.4.10 ROAZandROAZII[LSA]

ROAZIandROAZIIaretwocatamaran-shapedASVdevelopedbyLSA(LaboratóriodeSistemasAutónomos)at

ISEP (Instituto Superior de Engenharia do Porto). ROAZ I is a small fiberglass ASV designed to perform

environmentmonitoring,bathymetrymappingandsupport integratedoperationswithmultipleautonomous

vehicles in riverineandestuarineareas;ROAZ II is amedium-sizeHDPEASVdesigned foroceanoperations,

namelyhydrographicandbathymetrysurveying,andsupporttosecurity,searchandrescueoperations.Both

vehicles were used as a test beds for the development of navigation, control and collaborative operations

algorithms(Sonnenburg,2013;LSA30).

11.2.4.11 C-Enduro[ASV]

DevelopedbyASV, theC-Enduro is ahighendurance carbon fibre, self-righting catamaran. It isdesigned to

operateinallmarineenvironmentswithanenduranceofupto3monthsandhasgreatversatilityforarangeof

applications.Itssensorpayloadincludespassiveacousticmonitoring(PAM)andacommunicationlinkbetween

underwatervehicleandmonitoringstation.Forpowergeneration,theC-Enduroincorporatesdiesel,solar,wind

andwave powered options. ASView systemoffers threemodes of operation: direct, semi-autonomous and

autonomouscontrol(ASV31).

11.2.4.12 Viknes[NTNU&MarineRobotics]

TheViknes is anASVboatdesignedanddevelopedby theNorwegianUniversityof ScienceandTechnology

(NTNU) in conjunctionwithMarine Robotics. The vehicle is equippedwith inboard Yanmar 184 HPmotor,

boastingatopspeedofupto20kts(Sonnenburg,2013;Loe,2008;Viknes32].

11.2.4.13 Mariner[NTNU&MarineRobotics]

TheMariner is an automated inflatable-hull craft designed and developed by the Norwegian University of

ScienceandTechnology (NTNU) inconjunctionwithMarineRobotics.TheMariner isequippedwithadiesel

enginewithwaterjetpropulsion,withatopspeedofmorethan30knots.VehicleControlStation(VCS)isthe

interfaceusedformissioncontrolandmonitoringon-board.ThevehicleincludesaVHF/UHFradiolinkwitha

rangeofupto15km(MaritimeRobotics33).

30 http://www.lsa.isep.ipp.pt/roaz_home.html 31 http://asvglobal.com/product/c-enduro/ 32 http://www.viknes.no/ 33 http://www.maritimerobotics.com/systems/mariner/

143


DATE:July2016


11.2.4.14 C-Worker6[ASVUnmannedMarineSystems]

TheC-Worker isanunmannedrigid-hullcraftdevelopedbyASVUnmannedMarineSystemsforsecurityand

surveillanceasprimaryapplications,andisthemostversatilemodeloftheC-Workerfamily.Thisvehicleishighly

survivable,lightandrobustandischaracterisedbyamodularpayloadbayforeasyintegrationofstandardor

custompayloads.TheC-Worker6hasbeendeployedwithaSeichePAMarrayandremotemonitoringsystem,

makingseveralreal-timedetections.

11.2.4.15 C-Stat[ASVUnmannedMarineSystems]

TheC-Stat is a small single hull automated vehicle developed byASVUnmannedMarine Systems,with the

capability toremainonstationforextendedperiods.Amongthepossibleapplicationsarethepositioningof

subseaequipment,surfacetounderwatercommunications,oceanographicdatacollectionandsecuritysupport.

TheC-Statispoweredbydieselgeneratortochargethebattery,offeringanenduranceofupto4days.

11.2.4.16 ASV-300[C&CTechnologies&ASV]

The ASV-3000 high-endurance unmanned semi-submersible vehicle, akin to a small single-hull platform,

designedbyC&CTechnologiesandASV.PreviousdesignsofthistypeweretheORCA(OceanographicRemotely

Controlled Vehicle) fromNRL and theRMS (RemoteMinehunting System) from LockheedMartin (Wolking,

2011).

11.2.5 Autonomousunderwaterbuoyancygliders

11.2.5.1 SlocumG2electric[TeledyneWebb]

TheSlocumG2electricgliderismanufacturedbyTeledyneWebbResearch(TWR).Ithasa1000mdepthrating

witheitherdeepwaterorlittoralbuoyancyenginestooptimiseefficiency,amodularpayloadcapacitywitha

largevarietyofsensorsalreadyavailableonthemarketandaproventrackrecordindeliveryandoperation.It

hasahighenduranceof25toover365days,usingeitheralkalineor lithiumbatteries.Communicationisvia

Iridium/freewave or acoustic modem. A number of passive and active acoustic sensors have already been

integratedintotheSlocumgliderandusedtodetectmarinemammalsandfish(Table8).Ahybridversionofthis

glider exists which utilises a thruster with a collapsible propeller to provide momentum in sub-optimal

conditions.AllnewvehiclesareshippedwiththiscapabilityasastandardfeatureandexistingG2gliderscanbe

upgraded.Finally,althoughnotyetavailableascommercialofftheshelf(COTS)technology,athermalgliderhas

beendeveloped,wherepropulsionisfuelledbychangesintheoceantemperatureratherthanbatterypower

thatallowsgreaterlongevitytoanyonemission(3to5years).

144


DATE:July2016


11.2.5.2 Seaglider[Kongsberg]

DesignedattheUniversityofWashington(Eriksenetal.,2001),andmanufacturedbyKongsbergMaritime,the

Seaglider has a flooded aft section used to carry self-contained instruments. Steering of the vehicles is

undertakenbymovementoftheinternalmasses(e.g.batteries)bothforeandaftandforrotationaroundthe

axis.TheSeagliderispoweredbylithiumbatterieswithdeploymentsofupto10months,andcommunication

mayeitherbeusingIridiumorthroughacousticmodem.Anumberofpassiveandactiveacousticsensorshave

alreadybeenintegratedintotheSeagliderandusedtodetectmarinemammalsandfish(Table8).

11.2.5.3 Spray[ScrippsInstitutionofOceanography]

Designed at Scripps Institution of Oceanography (Sherman et al., 2001), the Spray gliderwas commercially

manufacturedbut isnotcurrently.TheSprayglider is similar to theSlocum,butwithamorehydrodynamic

shapeandhasbeenoptimisedforlong-rangeandadepthratingof1,500m.Turningisinitiatedbyrolling,like

theSeaglider.TheSprayuseslithiumbatteriesandmaximummissiondurationis~330days.

11.2.5.4 SeaExplorer[ACSA-ALCEN]

TheSeaExplorer,manufacturedbyACSA-ALCEN,hasbeendesignedtobepoweredbyrechargeablelithiumion

batteries, and to be faster (1 knot) than the other buoyancy gliders. It has a modular design including an

independent payload section located at the front of the vehicle. The SeaExplorer does not havewings and

communicationisbyIridium,radiooracousticmodem.Therechargeablebatteriesenableamaximummission

durationof~2months(1,200km)ononebatterycharge,refuellingtimeis20hoursandbatteryreplacement

every10years.SimilartotheSlocumhybrid,athrusteroptionisavailabletoenabletheglidertooperate in

powered AUV mode. A number of sensors including an acoustic recorder are already integrated into the

SeaExplorer.

11.2.5.5 eFòlaga[Graal-tech]

TheeFòlagaisahybridgliderconceptuallydesignedattheInteruniversityCentreofIntegratedSystemsofthe

MarineEnvironment(ISME),andengineeredbyGraal-tech(Caffaz,2009).Itisalight-weight,lowmaintenance

vehicleoptimisedforhighmanoeuvrabilitytoundertakecoastaloceanography.TheeFòlagahasbeendesigned

totransitbetweenstationsusingjetpropulsion(inpoweredAUVmode)andthenundertakesverticalprofiles

throughchangingitsballast(likeabuoyancyglider).ItusesaNiMhbattery,andhasanenduranceof6hours(at

maxpower),cantravelatspeedsupto4knotsandhasamaximumdepthof80m.Communicationisthrough

eitheraradiolinkoracousticmodem.TheeFòlagahasapayloadbaythatcanhostCOTSsensors,typicallyCTDs

145


DATE:July2016


havebeentestedsofar.However,trialsoftheeFòlagatotowapassiveacousticarrayhavebeenundertaken

formaritimesurveillance34.

11.2.5.6 Coastalglider[ExocetusDevelopmentLLC]

TheCoastalgliderwasdevelopedbyAlaskaNativeTechnologies,asamilitarizedgliderproduct (Imlachand

Mahr,2012).Thesegliders,oftentermedExocetus,weredesignedtooperateincoastalwaterswithhighdensity

gradientsandfastcurrentspeeds.Theyhaveabuoyancyengine7timeslargerthantheSlocumandSeaglider,

enablingthemtooperateoveralargerangeofwaterdensities.Alkalineorlithiumbatteriesenable14-60days

missionduration,andthebuoyancyglidercanaccommodatecurrentspeedsupto2knots.Communicationis

implementedthroughIridium,freewave(lineofsight),wifioracousticmodem.Anumberofdifferentsensors

havealreadybeenintegratedintotheCoastalglider,includingomni-directionalactiveacousticsensors(Imlach

andMahr,2012).OceanSonicssmarthydrophoneshavealsobeenintegratedintotheCoastalgliderandused

toprovideaproof-of-conceptvesselandfishinggearcollisionavoidancetechniqueusingthemeasuredradiated

noisefromthefishingvessels(Wood,2016).

11.2.6 Self-poweredsurfacevehiclesanddriftingsensorpackages

11.2.6.1 WaveGlidersSV2andSV3[LiquidRobotics]

OneofthemostestablishedvehiclescurrentlyavailablearetheLiquidRoboticsSV2andSV3Wavegliders.These

consistofasurfacefloat,aboutthesizeofasurfboard,andarackoffinssuspendedseveralmetresbelowthe

surface.Asthegliderliftsonawave,thefinstiltupwardsslightlyandsurgeforwardthroughthewater.When

thewaverelaxes,thefinstiltbacktothehorizontalandtheweightofthestructuresupportingthefinsagain

causesittosurgeforwards.Payloadscanbemountedinthefloatinghull,attachedtothesub-surfaceunitor

towedasternofthesub-unit. TheSV3 isaslightly largerversionoftheSV2andalsohasasmallelectrically

drivenpropelleronthesub-unitwhichcanbeusedincalmconditionswhennowavepowerisavailable(solar

power for this is generallymoreavailableon calmdays).Waveglidershavebeen successfully trialledwitha

numberofPAMsensorsformarinemammaldetection.

11.2.6.2 AutoNaut[MOST]

DevelopedbyMOST (AutonomousVessels) Ltd35,AutoNaut is another self-powered,wave-energypropelled

vehicle.Availableindifferentsizes:largerboatsshowhigherspeeds,greatercarryingcapacityandmorepower

forpayload.Thesevehiclesaregenuinelyself-poweredwithprimarypropulsionbydirectwavepropulsion(pitch

androll)usingWaveFoilTechnology.Solarpanelsharvestenergytopowerthesensorpackagewithbackup

34 http://www.sea-technology.com/features/2014/0214/8.php 35 http://www.autonautusv.com/

146


DATE:July2016


providedbybatteryandmethanolfuelcells.TheAutoNatdisplaysminimalnoise,issimpletodeploy/retrieve

andhasamissionenduranceofseveralmonths.TheAutonauthasperformedwellinseatrials,e.g.withPAM

sensors.

11.2.6.3 DASBRDriftingBuoy[NOAA]

TheDASBRbuoyhasbeendevelopedbyEmilyGriffithsandJayBarlowofNOAASouthWestFisheriesasalow

cost alternative to autonomous surface vehicles. Components, including an autonomous recorder and two

satellite trackers, can be purchased for around $5,000. Once deployed the buoys drift with the current.

Considerablecostscanthereforebeincurredintrackingdownandrecoveringthebuoys.

11.2.7 PAM

11.2.7.1 Decimus[SAInstrumentationLtd]

Thisisatthehighpowerendofthespectrumintermsofpowerconsumptionbutalsocontainsasophisticated

arrayofprocessingalgorithmswhichcanworkonuptofourchannelsata500kHzsamplerate.Ithassuccessfully

beenintegratedintobuoysandWavegliders.Datacanberelayedinnearreal-timeusingwirelessmodemsor

uploadeddailythroughcellphonenetworks.

11.2.7.2 DMON[WHOI]

DevelopedbyMarkJohnsonusingtechnologiesfromtheDTAG(Johnson,2003),thisultra-lowpowerdevicehas

successfullybeenprogrammedandintegratedintoaSlocumunderwaterglidertoautomaticallydetectbaleen

whales (Baumgartneretal.,2008;Baumgartneretal.,2013;Baumgartneretal.,2014b;Baumgartner,2014;

BaumgartnerandFratantoni,2008).TheDMONissubjecttoexportcontrolrestrictionsfromtheUS.

11.2.7.3 SoundTrapHFandSoundTrap4Channel[MarkJohnson]

Also developed by Mark Johnson, using similar technology to the DMON, this is another ultra-low power

recordingdevicewhichiscommerciallyavailablethroughOceanInstrumentsNewZealand.Thelatestsoftware

allows simultaneous recording at low frequencies combinedwith high frequency click detection,which can

extendrecordingdurationtoseveralmonths.Afour-channelversioniscurrentlyunderdevelopment.

11.2.7.4 SeicheReal-timeSystem[Seiche]

TheSeichewirelessPAMsystemisahighlyconfigurablesystemwhichcanbeutilisedfortruereal-timedata

transmission,withawirelesstransmissionrangeofupto10kms,typicallyusing2.4GHz/5GHzbands.Ithastwo

modestoenablereal-timemonitoring;Inthefirstmode,ananaloguetodigitalsamplingdeviceisinstalledwithin

theunitonthetransmittingplatform(e.g.anASV)andthefulldataset istransferredtoaPCatthereceiver

station (e.g. a support vessel) for processing in PAMGuard. The operator can then view and utilise the full

147


DATE:July2016


requiredfrequencyrangeformonitoringintruereal-time.Thisconfigurationrequireshigherbandwidthtoallow

receiptoffulldatasetatthereceiverstation.Inthesecondmode,anelectronicprocessingunitisinstalledon

thetransmittingplatformwithintheunitwheretheaudiosignalisprocessedthroughPAMGuard.Theoperator

atthereceivingstationhasfullcontrolandviewingaccessofincomingLFandHFdatawithinPAMGuarduser

interfaceviaa remote software link. This configuration runson lowerbandwidthandoffersa longer range.

Additionally,nodatalossissufferedshouldthelinkbelostasitispossibletorecorddataatthebaseunit.Both

configurationshavebeensuccessfullytrialledontheASVC-WorkerandC-Enduro,resultinginanumberoflive

detectionsofarangeofspecies.

11.2.7.5 WISPRBoard[OregonStateUniversity/Kongsberg]

TheWISPR(WidebandIntelligentSoundProcessor&Recorder)boardwasdevelopedandismanufacturedby

EmbeddedOceanSystems.TheWISPRsystemisamixedsignalmotherboardthatprocessesandlogsacoustic

signals.Itwasspecificallycreatedforapplicationswherebothspaceandpowerarelimitedsuchasfloats,gliders

andmoorings.

11.2.7.6 C-PODandC-POD-F[CheloniaLtd]

TheC-PODisawell-establishedclickloggingtechnologywidelyusedtostudyharbourporpoiseandothersmall

odontocete species around theworld. A simple processor logs key descriptors of each transient sound and

offlineanalysissoftwareclassifiestheseintoclicktrains.TheC-PODwastobetrialledontheC-Endurovehicle,

butthevehiclefailed.TheC-POD-F,whichincorporatesbetterprocessingandcollectsmoreaccuratedataon

eachloggedevent,willsoonbeavailable.

11.2.7.7 AutoDetectionBuoys[Cornell]

DevelopedforthedetectionofNorthAtlanticRightWhalesfrombuoysintheapproachestoBostonharbour,

software searches for specific sound types and sends a sampleof candidatedetections to shore forhuman

verification in near real time. To date, this has only been used onmoored buoys, butmay be suitable for

deploymentonautonomoussurfacevehicles.

11.2.7.8 A-tag[JapaneseScienceandTechnologyAgency]

TheA-tagisdevelopedbyTomomariAkamatsuoftheJapaneseScienceandTechnologyAgency,andis,similar

totheC-POD,aclickevent logger.Twochannelsareusedwhichenables ittocalculatebearingstodetected

sounds. It operates fully autonomously and is small enough that it could potentially be incorporated into

submarinegliders.

148


DATE:July2016


11.2.7.9 SM2M/SM3M[WildlifeAcoustics]

Thisisawell-establishedrecordingonlyunitwhichcansampleatupto384kHz.Theyareusuallyusedinmoored

systemsbuthavealsobeenusedindriftingbuoysystems.

11.2.7.10 AUSOMS-mini[AquaSound]

This isanotherrecordingonlysystem,samplingat44.1kHz. Ithasrelativelyshortrecordinglifetimebuthas

beendeployedongliders.LittleinformationisavailableoutthisproductsincethewebsiteisonlyinJapanese.

11.2.7.11 SDA14andSDA416[RTSYS]

DevelopedbyRTSYS,thesedevicescansampleatupto1MHz.Fourchannelsareavailableperboardandboards

canbeconnectedformorechannels.Thesystemscanmeasurenoise in1/3octavebandsandperformclick

eventdetection.TheyhavebeendeployedonsmallsizeAUV,Profilers,Surfacegliders,andBuoys.

11.2.8 AAM

11.2.8.1 ES853[Imagenex]

ThisisafullycommerciallyavailablesystemthathasalreadybeensuccessfullyintegratedintotheSlocumG2

gliderandSeaglider(ogive).Ithasafrequencyof120kHz,itsmaximumoperationdepthis1,000mandithasa

maximumdetectablerangeof100m.Itisprogrammabletohaveeitherafixed0.25Hzpingrate(inglidermode

forworkingonaSeaglider),ortobepolledbytheglider(Slocumscenario).Itsoperationalweightis1kginair

anditiscategorisedashavingnohazardclass.Theonlyinterferingfactortotheoperationofthissensoristhe

presenceofotheractiveacousticinstrumentsinthevicinity.

11.2.8.2 WBATandWBATmini[Kongsberg/Simrad]

ThesearebothfullycommerciallyavailablesystemsbuthavenotyetbeenintegratedintoanyAUVplatforms.

TheWBAT is a fully autonomousWideBand Autonomous Transceiver that stores data internally, operates

autonomously,andhasan internalbattery. It isastand-aloneunit thatcouldbedeployedonamooring,or

attachedtoanyROVorlargerAUV.TheWBAToperatesatthestandardSimradfrequenciesof38,70,120,200

and333kHz.TheWBATminiisbeingspecificallydesignedtobeintegratedintoverysmallvehiclessuchasa

SeagliderorsmallAUV.TheWBAToperatesusingbothnickelrechargeablebatteriesandlithiumbatterieswhich

affectsitshazardclass,whilsttheWBTminidoesnothavealistedhazardclass.Theonlyinterferingfactorto

theoperationofthesesensorsisthepresenceofotheracousticinstrumentsinthevicinity.

149


DATE:July2016


11.2.8.3 SontekADP[Sontek]

The Sontek ADP (Acoustic Doppler Profiler) is a fully commercially available system that has already been

successfullyintegratedintotheSCRIPPSSprayAUVglider(PowellandOhman,2015).Ithasamaximumprofiling

rangeof180mandoperatesat250kHz.

11.2.8.4 NortekADCP[Nortek]

The Nortek ADCP (Acoustic Doppler Current Profiler) is a fully commercially available system designed to

measurewatercolumncurrentsusingacousticDopplertechnology.A1MhzAquadoppprofilerhasalreadybeen

successfullyintegratedintotheSlocumG2glider(BaumgartnerandFratantoni,2008).TheAquadoppprofiler

weighslessthan3.5kg,hasamaximumprofilingrangeof100mandcanbeprogrammedtoeitherrecorddata

internallyortotransmitdatabyphoneorradiomodem.

11.2.8.5 DT-X-SUB[BioSonics]

ThisisafullycommerciallyavailablesystemandaversionofithasbeenintegratedintotheWaveglider(Greene

etal.2015).TheBioSonicsDT-X-SUB isa self-containedunit,designed forautonomousdeployments,witha

programmabledutycyclewhichenablesittobedeployedforlongtermstudies.Theunitcontrolsafullyfunction

split-beamechosounderandworkswithtransducersoperatingat38,70,120,200and420kHz.Environmental

performance limits include wind and wave conditions as this can impact the stabilisation of the towed

instruments.Theonlyotherinterferingfactortotheoperationofthissensoristhepresenceofotheracoustic

instrumentsinthevicinity.

11.2.8.6 AZFP[ASL]

The Acoustic Zooplankton Fish Profiler (AZFP) is currently commercially available as amoored system. It is

availablewith38,67.5,125,200,455,769and2,000kHztransducers,hasadepthratingupto1,000m.The

AZFPcanoperateinaninternallyrecordingmodeoritcanmakedataavailableinrealtime.Thecompanyoffers

compactAZFPpackagesforintegrationintoAUVandtowedbodies,butwearenotawareofthishavingbeen

testedyet.

11.2.8.7 ModularVR2CandVMT[Vemco]

TheVEMCOVR2Cacousticreceivers(frequency=69kHz)havebeenintegratedintoaSlocumG2glider(Haulsee

etal.2015).Theenvironmentalperformancelimitsofthesesystemsisthemaximumdepthatwhichtheycan

operate,whichis500mforthemodularVR2Cor1,000mfortheVMT.Theybothworkbydetectingthepings

fromdeployedacoustictagswhichcanbeattachedtoanyanimallargeenoughtobetagged.

150


DATE:July2016


11.2.8.8 Gemini720i[Tritech]

Thisisafullycommerciallyavailablemultibeamimagingsonarsystemwhichissuitablefordeploymentonavery

small ROVor AUV. The standard unit has environmental performance limits of 300mdepth and operating

temperatures between -10 and 35°C, however, a 4,000m depth ratedmodel is available for deeperwater

operations.ObjectdetectionandtrackingisavailablewiththeGeminiSeaTechsoftwarewhichallowstargetsto

beclassified.Thissystemhasbeenpreviouslyusedtodetectandclassifymarinemammaltargetsinreal-timeto

allowformitigationactionstobetakeninatimelymanner.

151


DATE:July2016


11.3 Comparisonmatrices

11.3.1 Listofplatforms

Table12.Autonomousplatformsincludedinthisreview,theirsystemclass(poweredandunpoweredASV,propellerandgliderAUV,lighter-than-airaircrafts(l-t-a)UASs,kitesandpoweredfixedwingUASs),systemname,manufacturerand/ordesigner/originatinguniversityaswellastheirtechnologyreadinesslevelasdefinedinTable10.Notethatcellswiththesamekindofinformationforsystemsofthesamemanufacturermaybemerged.

Systemclass

Systemname ManufacturerDesigner/

TechnologyreadinesslevelOriginatingUniversity

ASV

-po

wered

ASV-6300

ASV

Prototypesystem

C-Cat2 Firstofakindcommercialsystem

C-Enduro

FullcommercialapplicationC-Stat

C-Target3

C-Worker4 Firstofakindcommercialsystem

C-Worker6 Fullcommercialapplication

C-Worker7Firstofakindcommercialsystem

C-WorkerHydro

Delfim InstituteforSystemsandRobotics(Lisboa) AppliedResearch

Mariner NTNUandMaritimeRobotics Fullcommercialapplication

MeasuringDolphin UniversityofRostock(Germany) UniversityofRostock(MESSIN) Firstofakindcommercialsystem

ROAZI LaboratóriodeSistemasAutónomos Prototypesystem

152


DATE:July2016


Systemclass



ROAZII

RTSYSUSV RTSYS Demonstrationsystem

ASV

-un

powered

AutoNaut2

MOST(AutonomousVessels)Ltd FirstofakindcommercialsystemAutoNaut3

AutoNaut5

AutoNaut7

WavegliderSV2LiquidRobotics Fullcommercialapplication

WavegliderSV3

AUV-glider

ALBAC TokaiUniversity Smallscaleprototype

Coastalglider Exocetus AlaskaNativeTechnologies Fullcommercialapplication

Deepglider UniversityofWashington Demonstrationsystem

eFòlagaIII Graal-Tech IMEDEAInstitute Firstofakindcommercialsystem

LiberdadeXwing/Zray ONR Prototypesystem

Petrel TianjinUniversity,China n.p.

SeaBird KyushuInstituteofTechnology Prototypesystem

SeaExplorer ACSA Fullcommercialapplication

Seaglider(ogive) Kongsberg UniversityofWashington Fullcommercialapplication

SlcoumG2hybrid

TeledyneWebb Webb/WHOIFullcommercialapplication

SlocumG2glider

SlocumG2thermal Demonstrationsystem

Spray SCRIPPS Fullcommercialapplication

153


DATE:July2016


Systemclass



Sterneglider EcoleNationaleSuperioreD'IngenieursBrest n.p.

TONAI:TwilightOcean-ZonalNaturalResourcesandAnimalInvestigator

OsakaPrefectureUniversity,TaijiWhaleMuseum,Cetus

Basicresearch

AUV-prop

eller

A18-DECAGroup Fullcommercialapplication

A9-M

Bluefin-12D

BluefinRobotics Fullcommercialapplication

Bluefin-12S

Bluefin-21

Bluefin-9

Bluefin-9M

HUGIN1000(1000mversion)

KongsbergMaritime Fullcommercialapplication


HUGIN3000

HUGIN4500

MUNIN n.p.

REMUS100

Hydroid(KongsbergMaritime)+OSL(WHOI) FullcommercialapplicationREMUS3000

REMUS600

REMUS6000

RTSYSAUV RTSYS Demonstrationsystem

UAS

kite

SwanX1 FlyingRobotsSA Demonstrationsystem

154


DATE:July2016


Systemclass



UAS

l-t-a DesertStar10. AllsoppHelikitesLtd. Fullcommercialapplication

OceanEye MaritimeRoboticsAS Fullcommercialapplication

UAS-p

owered

-fixed

BRAMORC4EYE

C-Astral FullcommercialapplicationBramorgEO

BramorrTK

Fulmar Thales Fullcommercialapplication

Jump20 ArcturusUAV Fullcommercialapplication

PenguinB Uavfactory Fullcommercialapplication

ScanEagle Insitu Fullcommercialapplication

UX5 TrimbleNavigationLimited Fullcommercialapplication

155


DATE:July2016


11.3.2 Platformtechnicaldimensions

Table13.TechnicaldimensionsoftheautonomousplatformslistedinTable12.Givenaretheirmissiondurationandfactorslimitingthemissionduration,theirminimum,maximumandtypicalspeed,theirverticaloperationalrange,theirenvironmentalperformancelimits.Theiroperationalsizeandweightaswellaspacking/freightweightandsizearegivenandtherangetheycanbecontrolledwithin.Notethatcellswiththesamekindofinformationforsystemsofthesamemanufacturermaybemerged.Abbreviations:l-t-a:lighter-than-airaircrafts,NA:notapplicable,n.p.:notprovided,LWH:length,width,height.*unlessotherwisespecified

Missionduration Speed(knots)*Verticalrange

(m) Operational Packing/freight

Systemclass

Platforms MaxLimitingfactors

Min Max Typical Min MaxEnvperformancelimits

Size(m)*Weight(kg)

Size(m)*Weight(kg)

Controlrange(km):

ASV

-po

wered

ASV-6300 2-4days Fuelcapacity n.p. 6 n.p. NA NA n.p. ?x?x6.3 n.p. n.p. n.p. n.p.

C-Cat2

3-9hoursdependentonbatteryoption

Dieselcapacity

Stationkeeping(0)

5 3 NA NA

-10to45C(applicationspecificrangeavailable)

2.4x1.2x0.8LWH(0.2draft)

1002.4x1.2x0.8LWH

n.p.~2utilisingRFcomms

C-Enduro 90days

Energycapacity,sunlightanddiesel

7 3 NA NA4.2x2.4x2.8LWH(0.4draft)

500 5x2.6x2 n.p.

UnlimitedviaIridiumorothersatellitecomms/2-20+utilisingRFcomms

C-Stat 4days Fuelcapacity n.p. 3.7 n.p. NA NA n.p. ?x1.2x2.4 450 n.p. n.p. n.p.

C-Target3

Upto12hoursdependentonapplication

Dieselcapacity

3 25 6 NA NA

-10to45C(applicationspecificrangeavailable)

3.5x1.4x1.2LWH

3253.5x1.4x1.2LWH

n.p.2-20+utilisingRFcomms

C-Worker4 48hoursDieselcapacity

Stationkeeping(0)


7004.1x1.6x1.7LWH

n.p.UnlimitedviaIridiumorothersatellitecomms/2-20+utilisingRFcomms

C-Worker630days,Stationkeeping

Dieselcapacity


4,0006.3x2.2x2.2LWH

n.p.

C-Worker7Dependentonapplication

Dieselcapacity


5,3007.5x2.3x2.5LWH

n.p.

156


DATE:July2016




Systemclass



Size(m)*Weight(kg)

Size(m)*Weight(kg)

Controlrange(km):

C-WorkerHydro

6daysDieselcapacity

3 10 8 NA NA5.5x1.8x1.8LWH(0.9draft)

1,9005.5x1.8x1.8LWH

n.p.

Delfim n.p. Batterylife n.p. 5 n.p. NA NA n.p. ?x2x3.5 320 n.p. n.p. 80

Mariner50hours@5kts

Fuelcapacity n.p. >30 5 NA NA n.p.2.05x2.05x5.85

1,700 n.p. n.p. 15

MeasuringDolphin

3hours(batterypowered)or10hours(hybridpower)

Batterylife 0.5 4 n.p. NA NA n.p.3.3x1.8x1.5

250 n.p. n.p. n.p.

ROAZI n.p. Batterylife n.p. n.p. n.p. NA NA n.p.1.5x1x0.52

n.p. n.p. n.p. 3

ROAZII n.p. n.p. n.p. n.p. n.p. NA NA n.p.4.5x2.2x0.5

200 n.p. n.p. 3

RTSYSUSV 6hours Batterylife 1 6 3 0 0 Seastate 2x1x0.5 602.5x1.03x0.85

80 1

ASV

-un

powered

AutoNaut2 2weeks75WpPVpanel

0.25 2 1

surfacesensors;0.1

Towedsensors;20

Fullyoceancapable;stormproven

1x0.5x0.3 60 EuroPallet n.p. 5

AutoNaut3

3+months

Fuelformethanolfuelcell;anti-foulingsystem

0.5

3 2Towedsensors;50

2x0.4x05 1201.5x0.75x1.5

n.p.

oceanic;satellitecomms

AutoNaut5 4.5 3Towedsensors;100

3x0.8x1 2502.0x1.0x2.5

n.p.

AutoNaut7 6 4Towedsensors;200

4x1.1x1.5 4002.0x2.5x2.5

n.p.

WavegliderSV2

1year+Biofouling,Sunlightavailablity

0 2 1.25 NA NA

SurvivedSeaState8-9.Nopowerinflatcalm

~0.5x1x2x6depth

~100MultipleCrates,AirFreightOK

n.p. unlimited

157


DATE:July2016




Systemclass



Size(m)*Weight(kg)

Size(m)*Weight(kg)

Controlrange(km):

WavegliderSV3

0.3 2.5 1.75 NA NA

SurvivedSeaState8-9.Electricalpropellersinflatcalmifsufficientsunlight.

~0.5x1x3x5depth

~150 n.p.

AUV-glider

ALBAC1dive(30minutes)

Releaseofdiveweight

n.p. n.p.0.5-1m/s

n.p. 300

Currents,verticaldensitygradients

1.4(length)x0.24(diameter)x1.2(wingspan)

45 n.p. n.p. n.p.

Coastalglider

14days(alkaline),60days(lithium)

Batterylife n.p.1m/s

1m/s Surface 200 Currents2.87(length)x0.33(diameter)

109

2.44x0.85x0.72shippingcarton

n.p. unlimited

Deepglider 380days Batterylife n.p.0.45m/s

0.25m/s

Surface 6,000



62 n.p. n.p. unlimited

eFòlagaIII 6-8hours Batterylife n.p. n.p. 1-2m/s Surface 50-100


2.0(length)x0.16(diameter)

31 n.p. n.p. n.p.

LiberdadeXwing/Zray

6months Batterylife n.p. n.p.0.5-1.5m/s

n.p. 300


6.1(wingspan)

680 n.p. n.p. n.p.

Petrel n.p. n.p. n.p. n.p.

0.5m/s(2.0m/sthrust)

n.p. 500 n.p.


130 n.p. n.p. n.p.

SeaBird n.p. Batterylife n.p. n.p. n.p. n.p. 10Currents,vertical

0.45(length)x0.71

5 n.p. n.p. n.p.

158


DATE:July2016




Systemclass



Size(m)*Weight(kg)

Size(m)*Weight(kg)

Controlrange(km):

densitygradients

(width)x0.14(height)

SeaExplorer 2months Batterylife n.p. 1 1 Surface 700


2.0(length)x0.25m(diameter)x0.56(wingspan)x0.7(antenna)


Seaglider(ogive)

10months Batterylife n.p. n.p.0.25m/s

Surface 1000


1.8-2.0(length)x0.3(diameter)x1.0(wingspan)


SlcoumG2hybrid

360/50days(lithium/alkaline)

Batterylife

n.p.1m/s

0.25m/s

Surface200/1,000


1.5-2.15(length)x0.22(diameter)x1.2(wingspan)


SlocumG2glider

360/50days(lithium/alkaline)

n.p.0.38m/s

Surface200/1,000


SlocumG2thermal

3-5years n.p.0.38m/s

Surface 1,200 60 n.p. n.p. unlimited

Spray 330days Batterylife n.p. n.p.0.25m/s

n.p. 1,200



51

Boxdimensions1.6x0.6x0.6

92 unlimited

Sterneglider n.p. n.p. n.p. n.p. 1.3m/s n.p. n.p. n.p.4.5(length)x0.6(diameter

900 n.p. n.p. n.p.

TONAI n.p. n.p. n.p. n.p.0.2-0.5m/s

n.p. 60 n.p.1.65(length)x0.2(diameter)x

92 n.p. n.p. n.p.

159


DATE:July2016




Systemclass



Size(m)*Weight(kg)

Size(m)*Weight(kg)

Controlrange(km):

1.03(wingspan)

AUV-prop

eller

A18-D 24hours

Batterylife

n.p. 6 3 5 3,000 n.p.0.5x0.5x5.2

550-650 n.p. n.p. n.p.

A9-M

20hourswithtwoenergysections

n.p. 5 3 3 200 n.p.0.23x0.23x1.98

70 n.p. n.p. n.p.

Bluefin-12D

30hours@3ktswithstandardpayload

Batterylife

n.p. 5 n.p.

Surfacenavigation(~0)

1,500 n.p.0.32x0.32x4.32

260 n.p. n.p. n.p.

Bluefin-12S


n.p. 5 n.p. 200 n.p.0.32x0.32x3.77

213 n.p. n.p. n.p.

Bluefin-21


n.p. 4.5 n.p. 4,500 n.p.0.53x0.53x4.93

750 n.p. n.p. n.p.

Bluefin-9

12hours@3ktswithstandardpayload(SSS,cameraandprobes)

n.p. 5 n.p. 200 n.p.0.24x0.24x1.75

60.5 n.p. n.p. n.p.

Bluefin-9M

10hours@3ktswithstandardpayload(SSS,cameraandbackscattersensor)

n.p. 5 n.p. 300 n.p.0.24x0.24x2.5

70 n.p. n.p. n.p.

160


DATE:July2016




Systemclass



Size(m)*Weight(kg)

Size(m)*Weight(kg)

Controlrange(km):


24hours@4kts(withMBE,SSS,SBPandCTD)

Batterylife(dependsonspeed,sensorconfiguration,environmentandmissionprogram)

2 6 n.p.


1,000 n.p.0.75x0.75x4.5

850 n.p. n.p. n.p.



2 6 n.p. 3,000 n.p.0.75x0.75x4.7

850 n.p. n.p. n.p.

HUGIN3000

60hours@4kts(withMBE,SSS,SBPandCTD) Supplyoffuel

usedinthesemi-fuelcell

2 4 n.p. 3,000 n.p. 1x1x5.5 1,400 n.p. n.p. n.p.

HUGIN4500


2 4 n.p. 4500 n.p. 1x1x5.5 1,900 n.p. n.p. n.p.

MUNIN 12-24hours


n.p. 4.5 n.p.

1,500(available600version)

n.p.0.34x0.34x4

~300 n.p. n.p. n.p.

REMUS100

8hours@5kts;22hours@3kts


n.p. 4.5 3


100

Canbedeployedinroughweatherconditions

0.19x0.19x1.6

37 n.p. n.p. n.p.

REMUS3000

44hours@4kts(allsensorsoff)33hours@4kts(sonar

n.p. 4 3 3,000 n.p.0.36x0.36x3.7

335 n.p. n.p. n.p.

161


DATE:July2016




Systemclass



Size(m)*Weight(kg)

Size(m)*Weight(kg)

Controlrange(km):

+cameraon)

REMUS60045hours@4kts(allsensorson)

n.p. 5 4

600(available1,500configuration)

n.p.0.32x0.32x3.25

240 n.p. n.p. n.p.

REMUS6000

22hours@4kts

n.p. 5 n.p.

6,000(available4,000configuration)

n.p.0.71x0.71x3.84

862 n.p. n.p. n.p.

RTSYSAUV 20hours Batterylife 3 15 4to10 2 250Seastateforrecovery

150mm(diameter)x2m(long)

30Pelicase2.3

50 5

UAS-

kite

SwanX1 n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p.3.5x2.3x2.3

n.p. n.p. n.p. n.p.

UAS-l-t-a DesertStar

10.n.p. n.p. n.p. n.p. n.p. n.p. n.p.

Wind<72km/h

12x9 n.p. n.p. n.p. n.p.

OceanEye days n.p. n.p. n.p. 0 n.p. n.p. n.p. n.p. n.p.1.2x0.8x1.6

425 n.p.

UAS-p

owered

-fixed

BRAMORC4EYE

3hoursWind,airtemperature,sensortemperature,humidity,visibility,cloud,fog,rain,snow

32 58.32 58 30,000 5,000 Wind<15m/s0.02x2.30x0.96

4.5

Twoflightboxes.Catapultbox:125x45x30;Bramorbox:115x55x45

Catapultbox:28;Bramorbox:35

30

BramorgEO 2.5hours

BramorrTK 2.5hours

Fulmar 6-12hours n.p. n.p. n.p. 54 n.p. 800,000Wind<70km/h

n.p. 20 n.p. n.p. 70-90

Jump20 9-15hours n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p.

PenguinB 20+hours n.p. n.p. 69.98 43 n.p. n.p. n.p.?x3.3x2.27

21.5 n.p. 10 n.p.

162


DATE:July2016




Systemclass



Size(m)*Weight(kg)

Size(m)*Weight(kg)

Controlrange(km):

ScanEagle 24+hours n.p. n.p. 80 50-60 n.p. 5,944 n.p.?x3.11x1.55

22 n.p. 14-18 101.86

UX5 50minutesRain(onlytolerateslightrain),wind

n.p. n.p. 43 60,000 5,000 Wind<18m/s1.05x1.00x0.65

2.5 n.p. n.p. <5

163


DATE:July2016


11.3.3 Platformothertechnicaldetails

Table14.OthertechnicaldetailsoftheautonomousplatformslistedinTable12.Notethatcellswiththesamekindofinformationforsystemsofthesamemanufacturermaybemerged.Abbreviations:l-t-a:lighter-than-airaircrafts,NA:notapplicable,n.p.:notprovided,SSS:sidescansonar,CTD:conductivity,temperature,depth,IR:infrared,SVTP:sound,velocity,temperatureandtemperature,SUNA:SubmersibleUltravioletNitrateAnalyzer,LND,COTS:CommercialOff-The-Shelf,PAR:PhotosyntheticallyActiveRadiation,RiNKO:phosphorescentDOsensor.

Systemclass

Platforms Hazardclass Noiselevel Interferingfactors Typeofinterface AdditionalsensorsCTDsensordeployment

ASV

-po

wered

ASV-6300 n.p. n.p. n.p. n.p. MBE,SSS,SBE,SAS,CT,video n.p.

C-Cat2Class9:MiscellaneousDangerousGoods

Noisefromelectricpropellers

Propulsiondrives(Jet)Ethernet,Serial,ROS,proprietary

Sensorpackagesuserdefinable

Surface,winchindevelopment

C-EnduroGenerator,propulsiondrives

Winchandsurface

C-Stat n.p. Verylow n.p. n.p. Fullycustomizable n.p.C-Target3

None Low

Engine

Ethernet,Serial,ROS,proprietary

Sensorpackagesuserdefinable

NAC-Worker4 Propulsiondrive(Jet)

Surface,winchindevelopment

C-Worker6Generator,propulsiondrives

C-Worker7Generator,propulsiondrives

C-WorkerHydro Propulsiondrive

Delfim n.p. n.p. n.p. n.p.Single-beamimagingsonar(SBS),sidescansonar(SSS)

n.p.

Mariner n.p. n.p. n.p. n.p.EO/IRcamera,radar,oceanographicinstruments,SBES,MBES,sonar

n.p.

MeasuringDolphin n.p. n.p. n.p. n.p. Ultrasonicdepthfinder,ADCP n.p.ROAZI n.p. n.p. n.p. n.p. SSS,SonarAltimeter,CTD,Camera

WinchROAZII n.p. n.p. n.p. n.p.

SSS,SonarAltimeter,CTD,Camera,IRCamera

RTSYSUSV None n.p. n.p. Ethernet/serial n.p. Winchdeployed

ASV

-un

powered

AutoNaut2

None Silent None EthernetorSerialVerywiderangeofsensors;customerintegrationpossible

Hull/strutmounted;towedAutoNaut3

AutoNaut5 Hull/strutmounted;towed;winchedAutoNaut7

WavegliderSV2 n.p. VeryQuiet n.p. Serial 50+ Mounted

164


DATE:July2016


Systemclass


WavegliderSV3 n.p. n.p. Ethernet,SerialMounted,Winchcomingsoon

AUV-glider

ALBAC n.p. n.p. n.p. n.p. Temperature,velocity Mounted

Coastalglider n.p. n.p. n.p. n.p.

CTD/SVTP,acousticaltimeter,omnidirectionalsmarthydrophones,Wetlabswaterqualitysensors,RINKOdissolvedoxygen,SeabirdpumpedCTD,SUNAnitrate,LNDgammaray

n.p.

DeepgliderClass9:MiscellaneousDangerousGoods

n.p. n.p. n.p.SeabirdCTD,dissolvedoxygen,Wetlabsfluorometeropticalbackscatter

Mounted

eFòlagaIII Class8:Corrosives n.p. n.p. n.p. AnycustomorCOTSdevice Mounted

LiberdadeXwing/Zray

n.p. n.p. n.p. n.p.SPAWAR's32elementGliderTowedArraySystem(GTAS),on-boardpassiveacousticmonitoringsystem,DMON

n.p.

Petrel n.p. n.p. n.p. n.p. n.p. n.p.SeaBird n.p. n.p. n.p. n.p. n.p. n.p.

SeaExplorerClass9:MiscellaneousDangerousGoods

n.p. n.p. n.p.

CTD,dissolvedoxygen,turbidity,chlorophyll,CDOM,Phycobilins,HydrocarbonMinifluo-UVc,Methane,Metalstraces,Nitrate,acousticrecorder,altimeter…

Mounted

Seaglider(ogive)Class9:MiscellaneousDangerousGoods

n.p. n.p. n.p.

BiosphericalInstrumentsPAR,Aanderradissolvedoxygen,Seabirdelectronicsdissolvedoxygen,TurnerInstrumentsfluorometer/turbidimeter,Rocklandturbulencesensor,passiveacousticmonitoring,seabirdCTD,NortekacousticDopplercurrentprofiler,Wetlabsbackscattermeter/fluorometer,Controldissolvedoxygen,ImagenexES853

Mounted

SlocumG2hybrid

Class9:Miscellaneous

DangerousGoods(iflithiumbatteryused.Alsousesalkalinebatteries)

n.p. n.p. n.p. SeabirdpumpedCTD,Wetlabsfluorometer(chlorophylla),CDOM,turbidityandvolumescattering,Rinkoopticaloxygen,Vemcofishtrackerhydrophone,Satlanticmicro4channelirradianceandradiancesensors,BiosphericalInstrumentsPAR,SatlanticPar,Aanderaaoxygenoptode,ImagenexES853,TeledyneRDIADCP,SUNA

MountedSlocumG2glider n.p. n.p. n.p.

165


DATE:July2016


Systemclass


nutrientanalyser,RocklandscientificMicroRider

SlocumG2thermal n.p. n.p. n.p.

SeabirdpumpedCTD,Wetlabsfluorometer,chlorophylla,turbidityandvolumescattering,Rinkoopticaloxygen,Vemcofishtrackerhydrophone,Satlanticmicro4channelirradiance,andradiancesensors,BiosphericalInstrumentsPAR,SatlanticPar,Aanderaaoxygenoptode,ImagenexES853,TeledyneRDIADCP,SUNAnutrientanalyser,RocklandscientificMicroRider

SprayClass9:MiscellaneousDangerousGoods

n.p. n.p. n.p.

CTD,SeabirdCTD,SeapointOpticalBackscattersensor,Seapointchlorophyllfluorometer,TritechPA200acousticaltimeter,SontekArgonautAcousticDopplerCurrentProfiler,zooplanktoncamera

Mounted

Sterneglider n.p. n.p. n.p. n.p. n.p. n.p.

TONAI n.p. n.p. n.p. n.p.

RINKO-Profilerincludedepth,watertemperature,conductivity,salinity,dissolvedoxygen,chlorophyll-a,turbidity.A-Tag,camera

n.p.

AUV-prop

eller

A18-D n.p. n.p. n.p. n.p.Standardpayload:SSS,MBE,video,FLS(ForwardLookingSonar),CTDandenvironmentalsensors.

n.p.

A9-M n.p.

Lowmagneticandacousticsignature(STANAG1364compliant)

n.p. n.p.Standardpayload:SSS,video,SVP(CTDandenvironmentalsensorsonrequest)

n.p.

Bluefin-12D n.p. n.p. n.p. n.p.Variousavailablepayloads.Customizable

n.p.Bluefin-12S n.p. n.p. n.p. n.p. n.p.

Bluefin-21 n.p. n.p. n.p. n.p.Standardpayload:SSS,SBPandMBE.Customizable

n.p.

Bluefin-9 n.p. n.p. n.p. n.p.Standardpayload:dual-freq.SSS,camera,CTprobe,turbidityprobe.Customizable.

n.p.

Bluefin-9M n.p. n.p. n.p. n.p.Standardpayload:dual-freq.SSS,cameraandbackscattersensor.Customizable

n.p.

166


DATE:July2016


Systemclass



n.p. n.p. n.p. n.p.

Broadrangeofsensors,customizableBodymounted


n.p. n.p. n.p. n.p.

HUGIN3000 n.p. n.p. n.p. n.p.HUGIN4500 n.p. n.p. n.p. n.p.MUNIN n.p. n.p. n.p. n.p. NosemountedREMUS100 n.p. silent n.p. n.p.

Broadrangeofsensors,customizable

NosemountedREMUS3000 n.p. n.p. n.p. n.p.

REMUS600 n.p. n.p. n.p. n.p.Broadrangeofsensors,customizable(fullymodular)

REMUS6000 n.p. n.p. n.p. n.p. Broadrangeofsensors,customizable

RTSYSAUVClass9:MiscellaneousDangerousGoods

Lessthan100dBµPa

Speedandpayload Ethernet-WiFi Depth,Temperature n.p.

UASkite SwanX1Class3:FlammableLiquids

n.p. n.p. n.p. n.p. NA

UAS DesertStar10. Class2:Gases n.p. n.p. n.p. n.p. NAl-t-a OceanEye None n.p. n.p. n.p. n.p. NA

UAS-p

owered

-fixed

BRAMORC4EYE n.p.NoiselevelfortheBramorplatformis61.6dB(A),asmeasuredinSeptember2014.ThisiswellbelowtheEUnoisethresholds,whichis82dBinAustriaforexample.

Datacommunications

Radiofrequency,ifinthesamebandastheselecteddatatransmissionfrequency

Navigationlights,strobe,AN/PVS--7B/D,AN/PVS--14andAN/AVS--9compatibleIRbeacons,heatedpitot,multipleairvehiclecontrolfromsingleGCS,airpollution,radiation,hazardousandnon-hazardousgassensorADS--Btransponder

NA

BramorgEO n.p.

Include(butnotlimitedto):23.5MPRGBcamera,CIR,NDVI,4band,5bandor7bandMultispectralsensors,RikolaHyperspectralsensor,airpollution,radiation,hazardousandnon-hazardousgassensor(lasermassspectrometer),transponder,locatorbeacon.

NA

BramorrTK n.p.Include(butnotlimitedto):23.5MPRGBcamera,CIR,NDVI,4bandor5bandMultispectralsensors,locatorbeacon.

NA

Fulmar None n.p. n.p. n.p. n.p. NAJump20 None n.p. n.p. n.p. n.p. NAPenguinB None n.p. n.p. n.p. n.p. NA

167


DATE:July2016


Systemclass


ScanEagle None n.p. n.p. n.p. n.p. NAUX5 None n.p. n.p. n.p. n.p. NA

11.3.4 Platformcosts

Table15.CostdetailsoftheautonomousplatformslistedinTable12.Givenistheplatformpriceforpurchaseorrental,thecostsforoperation,maintenance,battery/fuelcostsandcostsforpilotingtelemetry.Notethatcellswiththesamekindofinformationforsystemsofthesamemanufacturermaybemerged.Pricesareforvehiclesonlywithoutsensors.Additionofsensorpackagescansignificantlyincreasetheprice.Abbreviations:l-t-a:lighter-than-airaircrafts,NA:notapplicable,n.p.:notprovided.

Price Costs

Systemclass

Platforms Purchase Rental Operation Maintenance Battery/Fuel Pilotingtelemetry

ASV

-po

wered

ASV-6300 n.p. n.p. n.p. n.p. n.p. n.p.C-Cat2 n.p. n.p. 2peopleforlaunchandrecovery,1

persononcontinuouswatch(usually4-6hourwatches)

Dependentonapplication

Rechargeablebatteries

DependentonapplicationC-Enduro n.p. n.p. 90Lforfulltanks

C-Target3 n.p. n.p. 40LC-Stat n.p. n.p. n.p. n.p. n.p. n.p.C-Worker4 n.p. n.p.

2peopleforlaunchandrecovery,1persononcontinuouswatch(usually4-6hourwatches)


100L

DependentonapplicationC-Worker6 n.p. n.p. 1100LfulltanksC-Worker7 n.p. n.p. 1200LC-WorkerHydro

n.p. n.p. 780Ltank

Delfim n.p. n.p. n.p. n.p. n.p. n.p.Mariner n.p. n.p. n.p. n.p. n.p. n.p.MeasuringDolphin

n.p. n.p. n.p. n.p. n.p. n.p.

ROAZI n.p. n.p. n.p. n.p. n.p. n.p.

ROAZII n.p. n.p. n.p. n.p. n.p. n.p.

RTSYSUSV $20,000-$50,000 $2,000-$5,000 $1,000-$2,000 $1,000-$2,000 $1,000-$2,000 $1,000-$2,000

ASV

-un

powered

AutoNaut2 $50,000-$100,000 $20,000-$50,000

Dependsonoperationalarea/requirement

$2000 $1000 UHF/WifiAutoNaut3

$100,000+ $50,000-$100,000$10,000

$2000SatellitecommsaccountAutoNaut5 $3000

AutoNaut7 $15,000 $5000

168


DATE:July2016


Price Costs

Systemclass


WavegliderSV2

n.p. n.p. n.p. n.p. NA n.p.

WavegliderSV3

n.p. n.p. n.p. n.p. NA n.p.

AUV-glider

ALBAC n.p. n.p. n.p. n.p. n.p. n.p.Coastalglider


DeepgliderNotyetacommercialproduct

Notyetacommercialproduct

NotyetacommercialproductNotyetacommercialproduct



eFòlagaIII n.p. n.p. n.p. n.p. n.p. n.p.LiberdadeXwing/Zray


Petrel n.p. n.p. n.p. n.p. n.p. n.p.SeaBird n.p. n.p. n.p. n.p. n.p. n.p.

SeaExplorer $100,000-$150,000 n.p. n.p. n.p.$0(rechargeablebatteries)

n.p.

Seaglider(ogive)

$100,000+ $50,000-$100,000 n.p.~$20,000(includingbatteries)

~$9,000/set ~$2,000-$3,000permonth

SlocumG2hybrid


SlocumG2glider


SlocumG2thermal


Spray

SpraywithCTD:$50,000-$100,000,withZooCamera$100,000-$150,000

Donotrent Pilotingis~$10/day/glider ~$10,000/year/glider

Batteryandrefurbishment~$12,000per4-5monthmission

Pilotingtelemetry~$2/daydatatransmissioncostscanbesignificantbutweusuallybringhomelargedatasetsonrecordedmedia.

Sterneglider


TONAI n.p. n.p. n.p. n.p. n.p. n.p.

AUV-

prop

eller A18-D n.p. n.p. n.p. n.p. n.p. n.p.A9-M n.p. n.p. n.p. n.p. n.p. n.p.Bluefin-12D n.p. n.p. n.p. n.p. n.p. n.p.Bluefin-12S n.p. n.p. n.p. n.p. n.p. n.p.

169


DATE:July2016


Price Costs

Systemclass


Bluefin-21 n.p. n.p. n.p. n.p. n.p. n.p.Bluefin-9 n.p. n.p. n.p. n.p. n.p. n.p.Bluefin-9M n.p. n.p. n.p. n.p. n.p. n.p.HUGIN1000(1000mversion)




HUGIN3000


HUGIN4500


MUNIN n.p. n.p. n.p. n.p. n.p. n.p.REMUS100 n.p. n.p. n.p. n.p. n.p. n.p.REMUS3000


REMUS600 n.p. n.p. n.p. n.p. n.p. n.p.REMUS6000


RTSYSAUV $100,000+ n.p. $2,000-$5,000 $2,000-$5,000 $1,000-$2,000 $1,000-$2,000

UAS

kite

SwanX1n.p. n.p. n.p. n.p. n.p. n.p.

UAS

l-t-a DesertStar

10.$50,000-$100,000 n.p. n.p. n.p. n.p. n.p.

OceanEye n.p. n.p. n.p. n.p. n.p. n.p.

UAS-p

owered

-fixed

BRAMORC4EYE

$100,000+startingprice(dependingonoptions)

TBDdependingonlocationandduration.Systemleaseispossibleviadifferentserviceproviders,notviathemanufacturer.

TBDdependingonlocationandduration

Completesparespartspackageis~$22,000for~1000flighthours

Includedinmantainancecosts.Onebatterypack(with2units)~$1,500

IncludedinOperationcostsBramorgEO

BramorrTK

Fulmar n.p. n.p. n.p. n.p. n.p. n.p.Jump20 n.p. n.p. n.p. n.p. n.p. n.p.PenguinB n.p. n.p. n.p. n.p. n.p. n.p.ScanEagle n.p. n.p. n.p. n.p. n.p. n.p.

170


DATE:July2016


Price Costs

Systemclass


UX5 n.p. n.p. n.p. n.p. n.p. n.p.

171


DATE:July2016


11.3.5 Platformsurveycapabilities

Table16.SurveycapabilitiesoftheautonomousplatformslistedinTable12.Givenisthecapabilityoftheplatformsfortracksetting,includingitsimplementationdetailsandlimitations,theirabilityforstationkeeping,autonomousdecisionmakingandclocksynchronisationincl.theirlimitations.Notethatcellswiththesamekindofinformationforsystemsofthesamemanufacturermaybemerged.Abbreviations:l-t-a:lighter-than-airaircrafts,NA:notapplicable,n.p.:notprovided.

Systemclass

PlatformsTracksetting

Implementationdetails

Trackkeepinglimitations

Selfcorrection

ExplanationStationkeeping

Limitationstostationkeeping

Autono-mousdecisionmaking

Decisionlimitations

Clocksynchro-nisation

Synchlimitations

ASV

-po

wered

ASV-6300

Yes Setwaypoint

n.p. n.p. n.p. n.p. n.p.

Yes

n.p. n.p. n.p.C-Cat2

Trackinglimitedbymaxspeedandendurancerequirements

Yes

SystemusescontrolsystemandGPSpositiontoprovideprecisetrackcorrection

Location


Thisisavailablethroughtestedandimplemented3rdpartysoftware

YesRequiressatellitevisibility

C-Enduro

C-Target3

C-Stat

No(stationkeepingbuoy)

None n.p. No NA Yes Seastate n.p. n.p. n.p.

C-Worker4

Yes Setwaypoint


Yes

SystemusescontrolsystemandGPSpositiontoprovideprecisetrackcorrection

Location


Thisisavailablethroughtestedandimplemented3rdpartysoftware

YesRequiressatellitevisibility

C-Worker6C-Worker7

C-WorkerHydro

Delfim Yes Setwaypoint n.p. Yes

Usespathfollowingtechniques,tomakethevehiclefollowaspecifictrackatspecifiedspeed.

n.p. n.p. Yes n.p. n.p. n.p.

Mariner Yes Setwaypoint n.p. n.p. n.p. No n.p. Yes n.p. n.p. n.p.

172


DATE:July2016


Systemclass




Selfcorrection




Decisionlimitations


Synchlimitations

MeasuringDolphin

Yes Setwaypoint n.p. n.p. n.p. n.p. n.p. Yes n.p. n.p. n.p.

ROAZIYes Setwaypoint

n.p. n.p. n.p. n.p. n.p.Yes

n.p. n.p. n.p.

ROAZII n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p.

RTSYSUSV Yes Setwaypoint Wave No n.p. Yes n.p. Yes n.p. Yes GPSsignal

ASV

-un

powered

AutoNaut2

Yes Setwaypoint

0.5ktcurrent

Yes Setwaypoint

Waypoints

<25mradiusfromwaypoint

Yes

Levelofautonomyprogrammedbeforemissioncommences

YesGPStimestamp;pre-missionsynchronisation

AutoNaut3 <2ktcurrent

AutoNaut5 <3ktcurrent Yes<35mradiusfromwaypoint

AutoNaut7 <4ktcurrentWaypoints

<45mradiusfromwaypoint

WavegliderSV2

Yes SetwaypointCurrents,Cloud,VesselTraffic

Yes Setwaypoint Yes<50mCEP

Yesn.p.

Yesn.p.

WavegliderSV3

<30mCEP NA n.p.

AUV-glider

ALBAC No n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p.Coastalglider

Yes Setwaypoint Current n.p. n.p. Yes Turnradius n.p. n.p. n.p. n.p.

Deepglider Yes Setwaypoint Currents Yes Other YesTurnradius,current

Yes n.p. n.p. n.p.

eFòlagaIII n.p. n.p. n.p.

Nounderwaternavigation/trackcorrection

n.p. n.p. n.p. n.p. n.p. n.p. n.p.

LiberdadeXwing/Zray

n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p.

Petrel n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p.SeaBird n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p.SeaExplorer n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p.

173


DATE:July2016


Systemclass




Selfcorrection




Decisionlimitations


Synchlimitations

Seaglider(ogive)

Yes Setwaypoint Currents Yes Other YesTurnradius,current

Yes n.p. n.p. n.p.

SlocumG2hybrid

Yes Setwaypoint Currents Yes Other Yes

Turnradius,current

Yes

n.p. n.p. n.p.

SlocumG2glider

Turnradius(7m,Davisetal2002),current

n.p. n.p. n.p.

SlocumG2thermal

Turnradius,current

n.p. n.p. n.p.

Spray Yes Setwaypoint Currents Yes Other YesTurnradius,current

Yes n.p.Seefootnote36

n.p.

Sterneglider


TONAI n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p.

AUV-prop

eller

A18-DYes Setwaypoint

n.p. n.p. n.p. n.p. n.p.Yes

n.p. n.p. n.p.A9-M n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p.

Bluefin-12D

Yes Setwaypoint


Yes

n.p. n.p. n.p.

Bluefin-12S n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p.Bluefin-21 n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p.Bluefin-9 n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p.Bluefin-9M n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p.HUGIN1000(1000mversion)

Yes Setwaypoint


Yes

n.p. n.p. n.p.


n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p.

HUGIN3000


36 Spray communicates with Iridium and GPS. While it is not programmed to do it, it could certainly "calibrate" its clock to milliseconds or better every 3-5 hours when it surfaces and share the calibration with a central site. We don't know much about its clock's stability.

174


DATE:July2016


Systemclass




Selfcorrection




Decisionlimitations


Synchlimitations

HUGIN4500


MUNIN n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p.REMUS100

Yes Setwaypoint

n.p. n.p. n.p. Yes Canholdstationeveninstrongcurrents(nolimitationsspecified)

Yes

Fullyauto-nomousbasedonpredefinedmissionplan

n.p. n.p.REMUS3000

n.p. n.p. n.p. Yes Yes n.p. n.p.

REMUS600 n.p. n.p. n.p. Yes Yes n.p. n.p.REMUS6000

n.p. n.p. n.p. Yes Yes n.p. n.p.

RTSYSAUV Yes SetwaypointDriftofsubseanavigation

No n.p. No NA YesRemotecontrolandmonitoring

Yes GPSorPTP/NTP

UAS

kite

SwanX1 YesManualpiloting

Wind YesManualpiloting

n.p. n.p. No n.p. n.p. n.p.

UAS

l-t-a DesertStar

10.no Other n.p. n.p. Other

Notintegrated

n.p. No n.p. n.p. n.p.

OceanEye no Other Wind No Other n.p. n.p. No n.p. n.p. n.p.

UAS-p

owered

-fixed

BRAMORC4EYE

Yes

Programmedwaypoints.Orautomatictargetfollowingfromoperatorselectiononvideoscreen.

Wind,airtemperature,sensortemperature,humidity,visibility,cloud,fog,rain,snow

Yes

Whenafailsafeisactivated,thesystemwillrespondaccordinglyandthenreturntoitslastmissionpositiontocontinueitsplannedflightmissionafterfailsafealertisinactive

n.p.

Timeofloiteringabovestationislimitedonlytobatterylifetime

No

NA

Yes

RecordedtimeoftheIRandRGBsensorsvideocanbesynchronizedwiththeGPSdata

BramorgEO

n.p. NA

BramorrTK n.p. NA

Fulmar Yes Setwaypoint n.p. Yes Setwaypoint n.p. n.p. No n.p. Yes

Controlcapacityforupto3UAVs,controltransfer

175


DATE:July2016


Systemclass




Selfcorrection




Decisionlimitations


Synchlimitations

betweenstations

Jump20 Yes Setwaypoint n.p. Yes Setwaypoint n.p. n.p. No n.p. n.p. n.p.PenguinB Yes Setwaypoint n.p. Yes Setwaypoint n.p. n.p. n.p. n.p. n.p. n.p.ScanEagle Yes Setwaypoint n.p. Yes Setwaypoint Location n.p. No NA n.p. n.p.UX5 Yes Setwaypoint Wind,rain Yes Setwaypoint No NA No NA n.p. n.p.

176


DATE:July2016


11.3.6 Platformoperationalinformation

Table17.OperationalinformationontheautonomousplatformslistedinTable12.Givenisthefueltype,failsafemodetype,transponder,detectandavoidcapacity,groundstationinterface,itslevelofautonomy,thepayloadcapacityintermsofpower,spaceandweightaswellasthedeploymentandrecoveryprocedure.Notethatcellswiththesamekindofinformationforsystemsofthesamemanufacturermaybemerged.Abbreviations:l-t-a:lighter-than-airaircrafts,NA:notapplicable,n.p.:notprovided.

Failsafemode Payloadcapacity Procedure

Systemclass

Platforms Fueltype Type MoreinfoTransponder,detectandavoidcapacity

Groundstationinterface

Levelofautono-my

PowerSpace

Weight(kg)

Deployment Recovery

ASV

-po

wered

ASV-6300 Diesel n.p. n.p. n.p. RFlink Waypoints n.p. n.p. n.p.ASV-6300SledL&Rsystem.Crane

ASV-6300SledL&Rsystem.Crane

C-Cat2 BatteryUsersettable

Failsafemodeisusersettable,dependingonrequirementsthesystemcanreturntowaypoint,orbitinplaceorstop

NoneIPradio,UHF

Waypoints

Upto250Wdependentonendurancerequirements

n.p. 50Slipway,Craneorhandlaunched

Slipway,Craneorhandrecovered

C-EnduroDieselElectrichybrid

n.p.

UsingAISsystemdetectandavoidavailable.radarandcameradetectandavoidindevelopment.

IPradio,Iridium,UHF

100Wcontinuous,250Wpeak

n.p. ~50Slipwayorcranelaunch

Slipwayorcranerecovery

C-Stat Diesel n.p. n.p.Optional(customizablepayload)

RF,satellite

Full(automatedstationkeeping)

n.p. n.p. 20 n.p. n.p.

C-Target3 PetrolUsersettable

Failsafemodeisusersettable,dependingonrequirementsthesystemcanreturntowaypoint,orbitinplaceorstop

NoneIPradio,UHF

Waypoints

Dependantonendurancerequirements

n.p.Dependentonapplication

Slipwayorcranelaunch

Slipwayorcranerecovery

C-Worker4 Diesel n.p. UsingAISsystemdetectandavoidavailable.radarandcameradetectandavoidindevelopment.

IPradio,Iridium,UHF

1000W

n.p. 80C-Worker6 Diesel

ElectricHybrid

n.p. n.p. 500

CraneLaunch CraneRecoveryC-Worker7 n.p. n.p. 500

C-WorkerHydro

Diesel n.p. 800W n.p. 120

Delfim None n.p. n.p. NoRFlink,80kmrange

Waypoints n.p. n.p. n.p.On-shoredeploymentorundocking

On-shorerecoveryordocking

177


DATE:July2016



Systemclass



Levelofautono-my

PowerSpace

Weight(kg)

Deployment Recovery

Mariner Diesel n.p. n.p. n.p. RF,Iridium Waypoints n.p.>1m3

n.p.On-shoredeploymentorundocking


MeasuringDolphin

None(2x400Wbatteriesforpropulsion,1x300Wbatteryforelectronicsandsensors)

n.p. n.p.

Collisionavoidancethroughforwardlookingechosounder

RDS(UHF/VHFradiolink)

Waypoints n.p. n.p. 100On-shoredeploymentorundocking


ROAZI

None(variablenumberof12V/3700Ahbatterypacks)

n.p. n.p.

Collisionavoidancethroughvideo

RF(WiFi802.11a/b/g)

Waypoints,targettracking

n.p. n.p. 50

On-shoredeploymentorundocking


ROAZII

None(4xAMG12V/56Ahbatterypacks)

n.p. n.p.RF(WiFi802.11a/b/g)

n.p. n.p. n.p.

RTSYSUSV Location n.p.Transponder,noavoidance

RF-WifiContinuouscontrol

n.p. n.p.50x40x30

5 Manual Manual

ASV

-un

powered

AutoNaut2 None

Endpoint

Homewaypointcanbesetpre-missionifrequired.

Radartransponder;AISClassBorC;autonomouscollisionavoidancebasedonAIStargetinformation

UHF;Wifi

Waypoints

5MJ

n.p.

20 Launchfromslipway/beachbyoneperson;launchfromsupportvesselwithdavit/crane

reverseoflaunch

AutoNaut3

Methanol,ifrequired

Acousticmodem,Iridium;UHF;wifi

50MJ 40

LARS(LaunchandRecoverySystem);operationsvan

AutoNaut5

Acousticmodem,Iridium;Inmarsat;UHF;wifi

175MJ 130

Launchfromslipway/beachbytwopeople;launchfromsupportvessel

178


DATE:July2016



Systemclass



Levelofautono-my

PowerSpace

Weight(kg)

Deployment Recovery

withdavit/crane

AutoNaut7Iridium;Inmarsat;UHF;wifi

300MJ 200

Launchfromslipway;launchfromsupportvesselwithdavit/crane

Reverseoflaunch

WavegliderSV2

NAKeeptrack

Willcontinueonlastprogrammedcourseorheadinguntilcommsarerestored

AISReceiver

Xbee,Acousticcommunications,iridium,WiFi,GatewayBuoy

Waypoints

Upto130W

n.p.

45+~2towed(neutrallybuoyant)

One-mandeployment

n.p.

WavegliderSV3

AISReceiver,fullyautonomousvehicleavoidance,optionalacousticreceiver

Acousticmodem,iridium,Wifi,Cell

Upto400W

60+~2towed(neutrallybuoyant)



AUV-glider

ALBAC Ni-Znbattery n.p. n.p. n.p. n.p. n.p. n.p. 3l n.p. n.p. n.p.

Coastalglider

Alkalineorlithiumbattery

n.p.

Emergencyrisemanoeuvre,emergencydive(heading)

altimeter,usehydrophoneforvesselavoidance

Iridium,ARGOS,FreewaveUHF,Wifi,acousticmodems

Waypoints

14MJ(alkaline),67MJ(lithium)or29.5MJ(rechargeable)

8L 5 Crane/davit Crane/davit

Deepglider

Lithiumsulfurylchloridebatteries

Location

Argos,bearing,location

Altitudesensorforbottomdetection.Icecopingalgorithm/behaviourcontrol

Iridium,acousticmodem

Waypoints17MJ(2*24V),4.4MJ(10VDC)

n.p. 25 Rib/crane Lasso

eFòlagaIIILeadacidorNiMhbatteries

n.p. n.p. n.p.

GPRS(cellular),wifi,GSM,24GHzradiolink

Waypointsandcontinuous

3.1MJ n.p. n.p. n.p. n.p.

179


DATE:July2016



Systemclass



Levelofautono-my

PowerSpace

Weight(kg)

Deployment Recovery

LiberdadeXwing/Zray


Petrel n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p. Crane/davit Crane/davitSeaBird Ni-MH n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p.

SeaExplorerRechargeablelithiumion

n.p.Autonomousdropweight,locationpinger,argos

n.p.Iridium,Radiofrequency,GPS

n.p. n.p. 9l 82people-rubberboat

2people-rubberboat

Seaglider(ogive)

Lithiumsulfurylchloridebatteries(optionalrechargeable)

Location

Argos,bearing,location


Iridium,acousticmodem

Waypoints 18MJ n.p.

4(Davisetal2002),25(Wood2009)

Rib/crane Lasso

SlocumG2hybrid

Alkalinebatteries,lithiumbatteries

Location

Dropweight,lastgaspmode,ARGOS


Iridium,Freewave900MHz(lineofsight),BenthosATM-900Modem

Waypoints

8MJ(Alkaline)

n.p.

5perpayloadbay(2payloadspossible)

Rib/craneNoseconerecoverySlocumG2

glider

Iridium,Freewave900MHz(lineofsight),BenthosATM-900Modem,ARGOS

8MJ(Alkaline),28MJ(needscheckingseeUNM2014_white)

n.p.

SlocumG2thermal

Iridium,Freewave900MHz(lineofsight),Benthos

6MJ n.p. 2

180


DATE:July2016



Systemclass



Levelofautono-my

PowerSpace

Weight(kg)

Deployment Recovery

ATM-900Modem

Spray

Lithiumsulfurylchloridebatteries

Location

Dropweight,surfacing,RFbeacon,acousticpinger

Altitudesensorforbottomdetection(Shermanetal2001)

Orbcomm,Iridiumsatellite

Waypoints 13MJ n.p. 3.5to51.8 Rib/crane n.p.

Sterneglider


TONAI n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p. 2 n.p. n.p.

AUV-prop

eller

A18-D

None

n.p. n.p.ObstacleAvoidanceSystem

Iridium,RF,WiFiandacoustic.Ethernet(onshore)

Waypoints

51.9MJ n.p. n.p. LARSorRHIB LARSorRHIB

A9-M n.p. EmergencypingerObstacleAvoidanceSystem(optional)

Iridium(optional),RF,WiFiandacoustic.Ethernet(onshore)

7.6MJperenergysection

n.p. n.p. Manual

Manual.LocalRemoteControlforsurfacerecovery

Bluefin-12D

None

n.p. n.p.

Acoustictrackingtransponder

RF,Iridiumandacoustic(remote).Ethernet(onshore)

Waypoints

27MJ(5x5.4MJLiPobatterypacks)

n.p. n.p.

ByA-frameorcrane

ByA-frameorcrane

Bluefin-12S n.p. n.p.

16.2MJ(3x5.4MJLiPobatterypacks)

n.p. n.p.

Bluefin-21 n.p. n.p.

48.6MJ(9x5.4MJLiPobatterypacks)

n.p. n.p.

Bluefin-9 n.p. n.p.RFandacoustic(remote).

5.4MJ(1xLiPobatterypack)

n.p. n.p. Manual Manual

181


DATE:July2016



Systemclass



Levelofautono-my

PowerSpace

Weight(kg)

Deployment Recovery

Ethernet(onshore)

Bluefin-9M n.p. n.p.

RF,Iridiumandacoustic(remote).Ethernet(onshore)

5.4MJ(1xLiPobatterypack)

n.p. n.p.Manualorbycrane

Manualorbycrane


None

Endpoint

Hominganddocking Collisionavoidance

withForwardLookingSonar(FLS)

Acousticmodem,iridium,WiFi

Waypoints

54MJ n.p. n.p.

LARS(safeuptoseastate5)

LARS(safeuptoseastate5)


54MJ n.p. n.p.

HUGIN3000

Typeusedinsemi-fuelcell(e.g.hydrogen)

162MJ n.p. n.p.

HUGIN4500

216MJ n.p. n.p.

MUNIN None n.p. 18MJ n.p. n.p.MinistingerorMUNINadapterforLARS

MinistingerorMUNINadapterforLARS

REMUS100

NoneEndpoint

Missionabort,hominganddocking

Abilitytodetect,locateandidentifyobjects.

Acousticcommunications,iridium,WiFi,GatewayBuoy Waypoints

3.6MJ n.p. n.p.One-mandeployment

n.p.

REMUS3000 Acoustic

modem,iridium,WiFi

n.p. n.p. n.p. LARS(LaunchandRecoverySystem);operationsvan

LARSREMUS600 18.7MJ n.p. n.p.REMUS6000

39.6MJ n.p. n.p.

RTSYSAUVLi-Ionbattery110

Endpoint

BuiltinFailuretree

n.p.WiFi-Iridium

Waypoints 3000Kj250mmlong

4Launchingramp

Handleorcradle

182


DATE:July2016



Systemclass



Levelofautono-my

PowerSpace

Weight(kg)

Deployment Recovery

220Vcharger

-Diameter140mm

UAS-kite

SwanX1 Gasoline None n.p. n.p. n.p.Continuouscontrol

n.p.

0.84x1.23x0.88

n.p. Runway n.p.

UAS-l-t-a DesertStar

10.NA None n.p. n.p. n.p.

Continuouscontrol

n.p. n.p. 30 Boattow Tetheredline

OceanEye NA None n.p. n.p. n.p.Continuouscontrol

n.p. n.p. n.p. Boattow Tetheredline

UAS-p

owered

-fixed

BRAMORC4EYE

Electric

n.p.

Multipleflightmodesandfailsafesavailable,user-configurable.LossofcommunicationsandGPSindicatorsandauralwarning.Lowbatteryindicationandauralwarning,criticalbatteryindicationandauralwarning.Terraindatasupportwithterrainnotificationandminimumterrainobstaclefailsafeoperation.Returntohomefeatures,

S-modetransponderisavailableasanoption

SurroundingRFcancauseinterference,ifthesameastheselectedtransmitter/receiverfrequency.Radiocommsareusuallybetween800-920MHzor902-928MHzradioformilitaryapplications(canbeadaptedaccording

Waypoints 7J

8cmhigh,83cmdiameter

0.3

Automaticcatapultlaunch(elasticorpneumaticcatapultoptions)

Parachuteautomaticlanding

BramorgEO

n.p.

89.8x172.8x77cmHLW

0.8

BramorrTK n.p.

89.8x172.8x77cmHLW

0.74

183


DATE:July2016



Systemclass



Levelofautono-my

PowerSpace

Weight(kg)

Deployment Recovery

emergencylandingfeatures.

tocountryTRA)

FulmarGasolineandheavyfuel

Location

n.p. n.p. RF Waypoints n.p. n.p. 8 Catapult Netlanding

Jump20 GasolineLocation

n.p. n.p. RF Waypoints n.p. n.p.30(inclfuel)

Verticaltakeoff Verticallanding

PenguinB Gasoline n.p. n.p. n.p. RF Waypoints n.p. n.p. 10Catapult,runwayorcarstoplaunch

n.p.

ScanEagle Gasoline n.p. n.p. n.p. RF Waypoints n.p. n.p. 3.4Automaticcatapultlaunch

AutomaticwithSkyHook®cable

UX5 Electric n.p. n.p. n.p. RF Waypoints n.p. n.p. n.p.Automaticcatapultlaunch

Bellylanding

184


DATE:July2016


11.3.7 Platformmanningrequirements

Table18.ManningrequirementsfortheautonomousplatformslistedinTable12.Givenisxx.Notethatcellswiththesamekindofinformationforsystemsofthesamemanufacturermaybemerged.Abbreviations:l-t-a:lighter-than-airaircrafts,NA:notapplicable,n.p.:notprovided.

Manningrequirements

Systemclass

Platforms PilotingMinnumberofpeopleforoperation

Vesselrequirements

Deployment Recovery Trainingneeds

ASV

-po

wered

ASV-6300Autonmous,supervisednavigationorremotecontrol

1 n.p. n.p. n.p. n.p.

C-Cat2

1continuouslyonshift

1Dependentonapplication

1-2people 1-2people

4daycourseC-Target3 1-22people 2people

C-Enduro 2-3

C-Stat Autonomous 1 n.p. n.p. n.p. n.p.

C-Worker4

1continuouslyonshift

2-3


2people 2people 4Daycourse

C-Worker6

3 3people 3peopleDependentonapplication

C-Worker7

C-WorkerHydro

Delfim Autnomous(basedonmissionplan) 1 None 2people 2people n.p.

Mariner Autonomous(basedonmissionplan) 1 n.p. n.p. n.p. n.p.

MeasuringDolphin

Autonomous(missionplan).AutopilotorRemoteControl(docking)

1 n.p. 2people 2people n.p.

ROAZI Autonmous,supervisednavigationorremotecontrol

1n.p.

2people 2peoplen.p.

ROAZII n.p. n.p.

RTSYSUSV n.p. 2 0 Trail Trail No

ASV

-un

powered

AutoNaut2 Dependsonoperationalmission/location;eg24/7monitoringinbusycoastalwaters,regularmonitoringwithautowarningsinquietoffshorewaters

1

Trainedtechnicianformaintenancebetweenmissions

1 1

Pilot&maintainertraining,3daysashorewith1dayatsea.

AutoNaut32

Possiblewith1,butusually2

SameaslaunchAutoNaut5

2AutoNaut7WavegliderSV2 Supervisionandmissioncontrol

throughVehicleIntefaceProgram(VIP),24hourpassive(alertbased)

1per20WaveGliders16mminimum,1tonneliftingcapability

2people+captain 2people+captainL&RTraining,3-dayclassWavegliderSV3 1per60WaveGliders 3people+captain 3people+captain

AU V- gli

de r ALBAC n.p. n.p. n.p. n.p. n.p. n.p.

185


DATE:July2016


Manningrequirements

Systemclass


Vesselrequirements


Coastalglider n.p. 2Rib/vesseltogettowaterdepth

2 2 n.p.

Deepglider 1-2people,occasionaltweaking n.p.Rib/vesseltogettowater>50m

1-2people 1-2people

Deployment/recoveryminimaltraining,pilotingrequirestrainingandcomputerskills

eFòlagaIII 1-2people,occasionaltweaking n.p. n.p. 1-2people 1-2people


LiberdadeXwing/Zray


Petrel n.p. n.p. n.p. n.p. n.p. n.p.

SeaBird n.p. 1 n.p. n.p. n.p. n.p.

SeaExplorer n.p. n.p. n.p. n.p. n.p. n.p.

Seaglider(ogive) 1-2people,occasionaltweaking n.p.Rib/vesseltogettowater>50m

1-2people 1-2people


SlocumG2hybrid

1-2people,occasionaltweaking

n.p.Shore/Rib/vesseltogettowater

1-2people 1-2people


SlocumG2glider n.p.Rib/vesseltogettowater>20mSlocumG2

thermaln.p.

Spray 1-2people,occasionaltweaking n.p.Rib/vesseltogettowater>20m

1-2people 1-2people


Sterneglider n.p. n.p. n.p. n.p. n.p. n.p.TONAI n.p. n.p. n.p. n.p. n.p. n.p.

186


DATE:July2016


Manningrequirements

Systemclass


Vesselrequirements


AUV-prop

eller

A18-D

Autonomous 1

n.p.WhateverLARSorRHIBmayrequire

WhateverLARSorRHIBmayrequire

n.p.

A9-M n.p.2people(2liftpointslocatedforeandaft)

2people(2liftpointslocatedforeandaft)

n.p.

Bluefin-12D

Autonomous.MonitoringthroughOperatorToolSuite

1

n.p. 1liftpointinthemiddleforcranerecovery

1liftpointinthemiddleforcranerecovery

n.p.Bluefin-12S n.p. n.p.Bluefin-21 n.p. n.p.

Bluefin-9

None



n.p.

Bluefin-9M

2people(2liftpointslocatedforeandaft).1extrareinforcedliftpointforcranerecovery

2people(2liftpointslocatedforeandaft).1extrareinforcedliftpointforcranerecovery

n.p.


SupervisionandmissioncontrolthroughHUGINOperatorSystem(HOS)

1Trainedtechnicianformaintenancebetweenmissions

1 1Pilot&maintainertraining

HUGIN1000(3000mversion)HUGIN3000HUGIN4500MUNIN

REMUS100

SupervisionandmissioncontrolthroughVehicleIntefaceProgram(VIP)

1Trainedtechnicianformaintenancebetweenmissions

1 1Pilot&maintainertraining

REMUS3000

REMUS600REMUS6000

RTSYSAUV Autonomousduringmisssion 2Nospecificrequirements

1personRigid-hulledinflatableboat

Yes

UAS-

kite SwanX1

1 1 NA n.p. n.p. UASpilotcertification

UAS

l-t-a DesertStar10. n.p. n.p. NA n.p. n.p. n.p.

OceanEye 1 1 NA n.p. n.p. UASpilotcertification

187


DATE:July2016


Manningrequirements

Systemclass


Vesselrequirements


UAS-p

owered

-fixed

BRAMORC4EYE1

1pilotand1payloadoperator(ideally)

NA 1person 1person 5daysBramorgEOBramorrTKFulmar n.p. 2 NA n.p. n.p. UASpilotcertificationJump20 n.p. n.p. NA n.p. n.p. UASpilotcertificationPenguinB n.p. 2 NA n.p. n.p. UASpilotcertificationScanEagle n.p. 2 NA n.p. n.p. UASpilotcertificationUX5 n.p. 2 NA n.p. n.p. UASpilotcertification

188


DATE:July2016


11.3.8 Listofsensor

Table19.Sensorsthatmaybeintegratedintoautonomousvehiclesandincludedinthisreview,theirsystemclass(AAM,PAM,Video),systemname,manufactureraswellastheirtechnologyreadinesslevelasdefinedinTable10).Notethatcellswiththesamekindofinformationforsystemsofthesamemanufacturermaybemerged.Abbreviations:AAM:activeacousticmonitoring,PAM:passiveacousticmonitoring.

Systemclass

Systemname Manufacturer Technologyreadinesslevel

AAM

Aquadopp Nortek Fullcommercialapplication

ADP Sontek Fullcommercialapplication

AZFP ASLBasicresearch

Fullcommerciallyapplicationasmooredsystem(forAUVs)

DT-XSUB BioSonics Fullcommercialapplication

ES853 Imagenex Fullcommercialapplication

Gemini720iTritech Fullcommercialapplication

Gemini720is

modularVR2CVemco Fullcommercialapplication

VMT

WBATKongsberg/Simrad

Fullcommercialapplication

WBTmini Prototypesystem

PAM

A-Tag MarineMicroTechnology Fullcommercialapplication

AUSOMS-miniBlack AquasoundInc. Fullcommercialapplication

C-POD-FChelonia

Smallscaleprototype

C-POD Demonstrationsystem

Cornell/AutoBuoys Cornell(PAM)/EOS(buoy) Ready

Decimus SAInstrumentationLtd Fullcommercialapplication

DMON WHOI AvailablefromWHOIn.p.

189


DATE:July2016


Systemclass

Systemname Manufacturer Technologyreadinesslevel

SDA14RTSYS

Demonstrationsystem

SDA416 Prototypesystem

Seicherealtimetransmissionsystem SeicheMeasurementsLtd n.p.

SM2/SM3 WildlifeAcoustics Fullcommercialapplication

SoundTrap4channelOceanInstruments

Prototypesystem

SoundTrapHF Fullcommercialapplication

WISPR EmbeddedOceanSystem n.p.

VIDEO

CM100UAVvision Fullcommercialapplication

CM202

DualImagerInsitu Fullcommercialapplication

EO900

OTUSU135HIGHDEFDST Fullcommercialapplication

OTUS-L205HIGHDEF

TASE310Cloudcaptechnology Fullcommercialapplication

TASE400HD

190


DATE:July2016


11.3.9 Sensortechnicaldetails

Table20.TechnicaldetailsofthesensorslistedinTable19.Givenaretheenvironmentalperformancelimitsofthesensors,theiroperationalaswellaspacking/freightsizeandweight,hazardclassasdefinedinTable11,anyfactorsinterferingwiththeoperationofthesensor,andthesensor’stypeofinterface.Notethatcellswiththesamekindofinformationforsystemsofthesamemanufacturermaybemerged.Abbreviations:AAM:activeacousticmonitoring,PAM:passiveacousticmonitoring,NA:notapplicable,n.p.:notprovided.

Operational Packing/freight

Systemclass

Systemname

Envperformancelimits Size(cm)* Weight(kg)* Size(cm)*Weight(kg)*

Hazardclass

Interferingfactors

Typeofinterface

AAM

Aquadopp n.p.47-71(height),7.5(diameter)

2,200 n.p. n.p. n.p. n.p. RS232,RS422

ADP n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p.

AZFP n.p.100(length),17(diameter)+transducersize

n.p. n.p. 50 n.p. n.p. RS422

DT-XSUB

Astableplatformisnecessaryfordatacollection.ThetowedbodyisisolatedfromthemechanicalsurgeoftheWaveGliderengineviaaspecifically-engineered,complianttowcable.However,extremewind/waveconditionsmayresultindestabilizationofthetowedbody.Thesystemispoweredviasolargeneratedelectricity,thusprolongeddarknesswillresultintemporaryshutdownuntilbatteriescanberecharged.

56(length),25(diameter)+transducerseither19or26(diameter)

Inwaterweight20lbs.(10,000g)

80x30x30 45 NoneOtheracousticinstruments

USBandethernetports

ES853 n.p.94(height),83(diameter)

1000 n.p. n.p. NoneOtheracousticinstruments

R232

Gemini720i

Standardis300m.availableina4,000mdepthratedmodelfordeep-wateroperations.Temperatureranges-10to35°C(operating)

n.p.1.2(300m),7.5(4000m)inwater

48x28x46(300m),53x50x50(4000m)

14(300m),20(4000m)

None NoneEthernet,VDSL,RS232,IsolatedTTL

191


DATE:July2016



Systemclass

Systemname


Hazardclass

Interferingfactors

Typeofinterface

Gemini720is

Standardis1,000m(aluminium).availableina4,000m(titanium)depthratedmodelfordeep-wateroperations.Temperatureranges-10to35°C(operating)

11x14x281.5(1000m),3.0(4000m)inwater

48x28x46(1000m),48x28x46(4000m)

14(1000m),17(4000m)

modularVR2C

500mdepthlimitpcb21(length),2(width)

99gair n.p. n.p. n.p. n.p. Serial

VMT 1000mdepthlimit18by3.5diameter

280gair n.p. n.p. n.p. n.p.

Optical-hasadedicatedreaderorcanbeBluetooth

WBAT n.p. n.p. n.p. n.p. n.p.

Class9:MiscellaneousDangerousGoods

Otheracousticinstruments

Self-contained/USBdownloadinterface

WBTmini n.p.10x10x15+wrapping

n.p. n.p. n.p. None n.p. n.p.

PAM

A-Tag 0-40Celcius,200m NA NA NA NA NA NA

A-tagisastand-alonesystemwithoutanyinterfacetotheoutside.

AUSOMS-miniBlack

0-40Celcius,1000m NA NA NA NA NA NA

AUSOMS-miniisastand-alonesystemwithoutanyinterfacetotheoutside.

C-POD-FAllmarinewatertemps;0-100m.Or0-2000m(DeepC-POD)

NA NA NA NA NA NA Noneyet

C-POD NA NA NA NA NA NANone,exceptviathirdparty

192


DATE:July2016



Systemclass

Systemname


Hazardclass

Interferingfactors

Typeofinterface

Cornell/AutoBuoys

NA NA NA NA NA NA NA Ethernet,serial

Decimus -10-60degc NA NA NA NA NA NASerial,Ethernet,WiFi

DMON n.p. NA NA NA NA NA NA n.p.

SDA140°C-80°C

NA NA NA NA NA NA Serial,ethernet,wifi,uhfSDA416 NA NA NA NA NA NA

Seicherealtimetranmissionsystem

n.p. NA NA NA NA NA NA n.p.

SM2/SM3Maximumdepthis150mfortheshallowversion,800mforthedeep-waterversion

NA NA NA NA NA NA None

SoundTrap4channel

Temp:-10to>50degC,depth:<1,000m

NA NA NA NA NA NA

RS232,RS485SoundTrapHF

NA NA NA NA NA NA

WISPR NA NA NA NA NA NA NA Ethernet,serial

VIDEO

CM100 n.p. 190x100 800g n.p. n.p. None n.p. n.p.

CM202 n.p. 295x190 3.5 n.p. n.p. None n.p. n.p.

DualImager


EO900 n.p. n.p. n.p. n.p. n.p. n.p. n.p. n.p.

OTUSU135HIGHDEF

n.p. 185x135 1.5 n.p. n.p. n.p. n.p. n.p.

OTUS-L205HIGHDEF

n.p. 256x205 2.6 n.p. n.p. n.p. n.p. n.p.

193


DATE:July2016



Systemclass

Systemname


Hazardclass

Interferingfactors

Typeofinterface

TASE310OperatingTemperatureRange:-20°Cto+60°C

Size:178x178x267mmTurretDiameter:178mm

3 n.p. n.p.None

n.p.ControlInterface:RS-232,CAN,Ethernet(withadaptor)

TASE400HD

3.5 n.p. n.p. n.p.

11.3.10 Sensorcosts

Table21.CostssensorslistedinTable19.Givenarethepurchaseandrentalprice,aswellasoperational,maintenance,battery/fuelaswellasintegrationcosts.Notethatcellswiththesamekindofinformationforsystemsofthesamemanufacturermaybemerged.Abbreviations:AAM:activeacousticmonitoring,PAM:passiveacousticmonitoring,NA:notapplicable,n.p.:notprovided.

Price Costs

Systemclass

Systemname Purchase Rental Operation Maintenance Battery/Fuel Integration

AAM

Aquadopp n.p. n.p. n.p. n.p. n.p. n.p.ADP n.p. n.p. n.p. n.p. n.p. n.p.

AZFPCAD$35,000-$80,000

3-monthrentalcostisinthe$10,000-$20,000range

Batterycostsareinthe$1,000-$2,000range

n.p. n.p. n.p.

DT-XSUB$50,000-$100,000

$5,000-$10,000 $1,000-$2,000 $1,000-$2,000 $1,000-$2,000 Bothpossible

ES853 $5,000-$10,000 NA IntegrationCost Nomovingparts Requires24vdc IntegrationCostGemini720i

$30,000-$50,000

$1,000-$2,000perweekNone;setupandrun/logdataonPC

None35Wpowerrequirement Software

DevelopmentKitavailableGemini720is 25Wpowerrequirement

modularVR2C$2,000-$5,000

NA n.p. n.p.Minimal-poweredoffgliderlessthan1milliamp

n.p.

VMT NA n.p. n.p. $350forannualrebatatfactory n.p.

WBAT $30,000-$50,000

n.p. n.p. n.p.BatteryiseitherlithiumorNiMh.Costis18,000NOKforlithium

n.p.

WBTmini n.p. n.p. n.p. n.p. n.p.

PAM A-Tag $5,000-$10,000 n.p. NA NA NA NA

AUSOMS-miniBlack

$2,000-$5,000 n.p. NA NA NA NA

194


DATE:July2016


Price Costs

Systemclass

Systemname Purchase Rental Operation Maintenance Battery/Fuel Integration

C-POD$2,000-$5,000 $1,000-$2,000

NA NA NA NAC-POD-F NA NA NA NACornell/AutoBuoys

TBD TBD NA NA NA NA

Decimus n.p. n.p. NA NA NA NADMON n.p. n.p. NA NA NA NASDA14 $2,000-$5,000 n.p. NA NA NA NASDA416 $5,000-$10,000 n.p. NA NA NA NASeicherealtimetranmissionsystem

n.p. n.p. NA NA NA NA

SM2/SM3 $5,000-$10,000 n.p. NA NA NA NASoundTrap4channel $2,000-$5,000

n.p. NA NA NA NA

SoundTrapHF n.p. NA NA NA NAWISPR n.p. n.p. NA NA NA NA

VIDEO

CM100 n.p. n.p. n.p. n.p. n.p. IntegratedCM202 n.p. n.p. n.p. n.p. n.p. IntegratedDualImager n.p. n.p. n.p. n.p. n.p. IntegratedEO900 n.p. n.p. n.p. n.p. n.p. IntegratedOTUSU135HIGHDEF

n.p. n.p. n.p. n.p. n.p. Integrated

OTUS-L205HIGHDEF

n.p. n.p. n.p. n.p. n.p. Integrated

TASE310 n.p. n.p. n.p. n.p. n.p. IntegratedTASE400HD n.p. n.p. n.p. n.p. n.p. Bothpossible

195


DATE:July2016


11.3.11 SensorsurveycapabilitiesI

Table22.FirstsetofsurveycapabilitycriteriaforsensorslistedinTable19.Givenisthedatatypecollected,dataprocessingdetails,ifspeciesidentificationispossibleandifso,whichtype,andifthebearing,directand/orhorizontalrangetotheanimalisestimablewiththesensordata.Notethatcellswiththesamekindofinformationforsystemsofthesamemanufacturermaybemerged.Abbreviations:AAM:activeacousticmonitoring,PAM:passiveacousticmonitoring,NA:notapplicable,n.p.:notprovided.

EstimabletoanimalSystemclass

Systemname

Collecteddatatype

ProcessesSpeciesidentification

Speciestype BearingDirectrange

Horizontalrange

AAM

Aquadopp Raw n.p. Notproven Fish/zooplankton Yes Yes NoADP Raw n.p. Notproven Fish/zooplankton Yes Yes No

AZFP Raw n.p.Ifmultifrequency,possible

Fish/zooplankton Yes Yes No

DT-XSUB Raw

On-boardprocessinggeneratesdatasummarieswhichcanbetransmittedandviewedassimplifiedechograms.Itisalsopossibletogeneratealertstriggeredbyacousticevents(detectionoftarget(s)havingTSexceedingaspecificthreshold)

Ifmultifrequency,possible

Fish/zooplankton Yes Yes No

ES853 Raw n.p. Notproven Fish/zooplankton Yes Yes NoGemini720i

RawGeminiformatimagedatalogged

SonardataiscollectedinaproprietaryTritechformat,whichcanbereplayedatthesamequalityitwasrecorded.ThirdpartysoftwarewouldneedaTritechECDfileconvertertoaccessthedatainthefiles.TheGeminiSeaTecSystemutilisestheindustrystandardGeminihardware,combinedwithGeminiSeaTecsoftware(anadaptionofTritech’sstandaloneGeminisoftware).SeaTecsoftwareprovidesintuitiveobjectdetectionandtargettrackingcapabilities.Targetsareclassifiedbasedontheprobabilityofthembeingaspecific,predeterminedtype.Forexample,marinemammals,pipesordebris.

YesMarinemammalautomaticdetection,rudimentaryfishdetection

Yes Yes YesGemini720is

modularVR2C Tagdetections AcousticpingsaredecodedintotagIDs Yes

Anyaquaticspecies-largeenoughtoholdanacoustictag

No No NoVMTWBAT

Rawn.p. Ifwideband,

probableFish/zooplankton Yes Yes No

WBTmini n.p.

PAM

A-TagSummaryclickinformation

Akamatsuetal(2005),MarineTechnologySocietyJournal39(2),3-9.

Smallodontocetes,snappingshrimps

IdentificationofDelphinidadandPhocoenidaefamily

Yes No No

196


DATE:July2016



Systemname

Collecteddatatype



Horizontalrange

AUSOMS-miniBlack

RawrecordingofLinearPCM,MP3andWMAformats

CommonlyusedcompressionformatsCetaceans,fish,shrimps

Dependsonoff-lineanalysis

No No No

C-POD

Descriptorsofultrasonictonesandclicksasdetectiondataandasinputtotraindetectionprocess.

Selectionofultrasonicclicksandtones;estimationofkeydescriptorsoftheseforstorage.

Allodontoceteclicks>20kHz

Somespeciesgroups Yes Yes No

C-POD-F

Descriptorsofultrasonictonesandclicksasdetectiondataandasinputtotraindetectionprocess.Waveformsofselectedclicksintrainstoaidspeciesidentification.

Selectionofultrasonicclicksandtones;estimationofkeyparametersoftheseforstorage.Traindetectiontoidentify'specimen'cetaceanclicksforstorageoffullwaveform.

Allodontoceteclicks>17kHz

Cornell/AutoBuoys

Compressedrawdata(FLAC),detections

n.p.

NRW,BHW,anyspeciesforwhichatrustedalgorithmexists

NRW,BHW,anyspeciesforwhichatrustedalgorithmexists

Yes,withanarrayofunits



DecimusRawrecordings&binarydata

Variousdetectors/noisemonitorsavailable

Harbourporpoise,Bottlenosedolphins,Spermwhales,Baleenwhales,Belugawhales,Rightwhales

Harbourporpoise,Bottlenosedolphins,Spermwhales,Baleenwhales,Belugawhales,Rightwhales

Yes No No

DMON n.p. n.p. n.p. n.p. n.p. n.p. n.p.SDA14 Raw(24bits.Wav),

.txtlogfiles;pre-computedinformation,eventdetection

Thirdoctavebands,clicklikeeventdetection,advancedspecificprocessingonrequest

Porpoise,dolphins(requirespostvisualidentification),whales

Porpoise,whales Yes Yes YesSDA416



197


DATE:July2016



Systemname

Collecteddatatype



Horizontalrange

SM2/SM3

Rawrecordingsarestoredeitherinuncompressed.wavfileformatorwithWAC0losslesscompressionwhichsavesapproximately50%oncardspace.

n.p.

Datahasbeenusedtorecordallmarinemammaltypesincludingwhales,dolphins,porpoises,manatees,etc

n.p. No No No

SoundTrap4channel Rawand/or

detectionsclickdetector-energyinband Toothedwhales

n.p. Yes Yes Yes

SoundTrapHF

n.p. No No No

WISPRCompressedrawdata(FLAC),detections

n.p.Anyspeciesforwhichatrustedalgorithmexists

Anyspeciesforwhichatrustedalgorithmexists




VIDEO

CM100 Raw n.p. n.p. n.p.Yes Yes Yes

CM202 Raw n.p. n.p. n.p.DualImager

Raw n.p. n.p. n.p.Yes Yes Yes

EO900 Raw n.p. n.p. n.p.OTUSU135HIGHDEF

Raw n.p. n.p. n.p.Yes Yes Yes

OTUS-L205HIGHDEF

Raw n.p. n.p. n.p.

TASE310 Raw n.p. n.p. n.p.Yes Yes YesTASE

400HDRaw n.p. n.p. n.p.

198


DATE:July2016


11.3.12 SensorsurveycapabilitiesII

Table23.SecondsetofsurveycapabilitycriteriaforsensorslistedinTable19.Givenisthereal-timedatatransmissioncapabilitiesincl.detailswatistransmitted,theabilityforclocksynchronisationanditslimitationsandforacousticsensors,ifreceivinglevelsofreceivedsignalsisestimableaswellasambientnoiselevels,andifCTDdataareobtainedsimultaneously.Notethatcellswiththesamekindofinformationforsystemsofthesamemanufacturermaybemerged.Abbreviations:AAM:activeacousticmonitoring,PAM:passiveacousticmonitoring,NA:notapplicable,n.p.:notprovided.

Realtimedatatransmission Clocksynchronisation Acousticsensorspecific

Systemclass

Systemname Possible Whatistransmitted Possible Limitations

EstimationCTDdataReceiving

level

Ambientnoiselevel

AAM

Aquadopp PartlyApingcanbetransmitted,butmultiplepingscollectedonasingledive

n.p. n.p. n.p. n.p. no

ADP PartlyApingcanbetransmitted,butmultiplepingscollectedonasingledive


AZFP PartlyApingcanbetransmitted,butmultiplepingscollectedonasingledive


DT-XSUB PartlyApingcanbetransmitted,butmultiplepingscollectedonasingledive

Yes None n.p. n.p. no

ES853 PartlyApingcanbetransmitted,butmultiplepingscollectedonasingledive

Yes n.p. n.p. no no

Gemini720i

Yes

CurrentlyalarmscanbesentoverCOM(serial)orparallelports.TritechcanadaptthesoftwareasrequiredandiscurrentlydoingthisforacustomerthatrequireswarningstobesentasHTTPPOSTcommandsovertheinternet,whichresultinamobilephoneTEXTmessagewhenatargetofaspecifictypeisifentifiedwithhighprobability.

Thesoftwareisdevelopedinhouseandcanbeadaptedasrequired.Basiclevelsofsynchronisationareachievedorwithpurchaseofextrahardwarehighperformanceclocksynchronisationisavailable.

ThesoftwareexpectsaninputPPSandZDAtimestring.

No No

Additionaldatacouldbeloggedifrequiredwithsoftwareadaptation

Gemini720is

modularVR2C Yes DetectedtagIDs Yes

Ifdetectiondataistransmittedinrealtimetoon-boardcomputer,thetimestamppfthedetectioncanbecorrectedbythecurrenttimeoftheon-boardcomputer

n.p.No No

VMT No n.p. No n.p. n.p.

199


DATE:July2016



Systemclass


EstimationCTDdata

Receivinglevel

Ambientnoiselevel

WBATPartly

Apingcanbetransmitted,butmultiplepingscollectedonasingledive

n.p. n.p. n.p. n.p. n.p.WBTmini n.p. n.p. n.p. n.p. n.p.

PAM

A-Tag No NA No NA yes No n.p.AUSOMS-miniBlack

No NA No NA Yes Yes n.p.

C-POD No NA No NAYes Yes

n.p.C-POD-F yes Samedataasstoredi.e.compressed Yes Wiredlinkbetweenunits n.p.

Cornell/AutoBuoys

Yes,withiridium/cellantenna

Detections/audioclips Yes GPSYes,withcalibratedsensor

Yes,withcalibratedsensor

n.p.

Decimus yes Binarydetectiondata/noise Yes indevelopment Yes Yes n.p.DMON n.p. n.p. n.p. n.p. n.p. n.p. n.p.SDA14

YesWavfiles,precomputednoisedata Yes GPS/PPSonly Yes Yes n.p.

SDA416 n.p. Yes GPS/PPS,NTP,PTP Yes Yes n.p.Seicherealtimetranmissionsystem


SM2/SM3 No NA Yes

UsetheSM3GPStosynchtheclockoftheSM3Mandthendetachfordeployment

YesDatahasbeenused

n.p.

SoundTrap4channel

n.p.Endusertoimplement

Yes GPSinput Yes Yes n.p.

SoundTrapHF n.p. Yes GPSinput Yes Yes n.p.

WISPRYes,withiridium/cellantenna

Detections/audioclips Yes GPSYes,withcalibratedsensor

Yes,withcalibratedsensor

n.p.

VIDEO

CM100Yes

n.p. n.p. n.p. NA NA NACM202 n.p. n.p. n.p. NA NA NADualImager

Yesn.p. n.p. n.p. NA NA NA

EO900 n.p. n.p. n.p. NA NA NAOTUSU135HIGHDEF

Yes n.p. n.p. n.p. NA NA NA

200


DATE:July2016



Systemclass


EstimationCTDdata

Receivinglevel

Ambientnoiselevel

OTUS-L205HIGHDEF

n.p. n.p. n.p. NA NA NA

TASE310Yes

n.p. n.p. n.p. NA NA NATASE400HD n.p. n.p. n.p. NA NA NA

11.3.13 Sensoroperationalrequirements

Table24.OperationalrequirementsofthesensorslistedinTable19.Givenistheirlevelofautonomy,deploymentandrecoveryprocedure,theirpayloadpower,andinternalaswellasexternalpayloadcapacityintermsofspaceandweight.Notethatcellswiththesamekindofinformationforsystemsofthesamemanufacturermaybemerged.Abbreviations:AAM:activeacousticmonitoring,PAM:passiveacousticmonitoring,NA:notapplicable,n.p.:notprovided.

Procedure Internalpayload

capacityExternalpayloadcapacity

Systemclass

Systemname

Levelofautonomy Deployment Recovery PayloadpowerSpace(cm)*

Weight(kg)*

Space(cm)* Weight(kg)*

AAM

Aquadopp Self-logging,programmedintervalsMountedonglider

Mountedonglider

0.5-1.5Wat1Hz0.47-0.71length0.08mdiameter

2.2-3.4inair

n.p. 2.2-3.4inair

ADP n.p.Mountedonglider

Mountedonglider

200-1000mW n.p. n.p. n.p. n.p.

AZFP n.p.Mountedonglider/AUV

MountedongliderAUV

n.p. n.p. n.p.0.17mdiameter,1mlength

n.p.

DT-XSUB NAMountedonglider

Mountedonglider

11-24volts NA NA NA NA

ES853Self-logging,fixedpingrate,polledbyAUV/glider

Mountedonglider

Mountedonglider

<250mW n.p. 1 n.p. 1

Gemini720i Thesonarisprovidedasaplugandplaydevice.Targetidentificationandclassificationwithvariablelevelsofprobabilityiseasilyenabledthroughuserselection.

Deploymentonasubseaframeoronapole

Retrievalusingaliftingwire;noexcessivestrainoncable

35Wpowerrequirement n.p. n.p. n.p. n.p.

Gemini720is

25Wpowerrequirement n.p. n.p. n.p. n.p.

201


DATE:July2016




Systemclass

Systemname


Weight(kg)*


modularVR2C

n.p.Embeddedinglider

Realtimetransmissionofdatathruglidercomms

24to180milliwattsPcb21(length),2(width)

99gair n.p. n.p.

VMT n.p.Mountedonglider

Mountedonglider

on-boardCcellbattery-factoryreplaceable

n.p. n.p. n.p. n.p.

WBAT Self-logging,fixedpingrateMountedonAUV

MountedonAUV n.p. n.p. n.p. n.p. n.p.

WBTmini n.p.Mountedonglider

Mountedonglider


PAM

A-Tag Fullyautonomousaftersettingup NA NA 60mW 5x5x54 0.6inair 5x5x54 0.6inair

AUSOMS-miniBlack

Fullyautonomousaftersettingup NA NA 75mW

5.1x5.1x19.3(Diameter5.1,Length19.3,CarbonFiberCase)

0.6inair(0.08inwater)

5.1x5.1x19.3(Diameter5.1,Length19.3CarbonFiberCase)

0.6inair(0.08inwater)

C-POD400yearsofdataautomaticallyprocessin5days,SAMBAHproject

NA NA 44mWcylinder540long/80OD

n.p.90diameter,606length

n.p.

C-POD-F High NA NA 60mW,lessifquiet n.p.91diameter,640length

n.p.

Cornell/AutoBuoys

n.p. NA NA <2Watts NA NACylindricalDelrinhousing(~40x20)

5

Decimus High NA NA2/3fordevice,commsdependant

20x20x10approx.

120x20x10approx.

n.p.

DMON n.p. NA NA n.p. n.p. n.p. n.p. n.p.

SDA14

Autonomous

NA NA600Mw-1.8W(multichannels)

14x6.5 200gDependsifhousingrequired,fromrecordertobuoyorothercarrier

VariableSDA416 NA NA n.p. 14x6.6 300g

Seicherealtimetransmissionsystem

n.p. NA NA n.p. n.p. n.p. n.p. n.p.

202


DATE:July2016




Systemclass

Systemname


Weight(kg)*


SM2/SM3Canbeeasilyprogrammedforautonomousautomatedrecordings.

NA NA

Powerusagevariesbysamplerate.Whenidle,theSM3Mconsumesaround0.5mW.Dependingonaccessories,samplerates,compression,andothervariables,theSM3Mcanuseaslittleas500mWofpowerwhenrecording.

Dimesionsofelectronics:22”longx5.25”widex4.25”high

2.95lbs

Length/Heightofhousing:90.9+/-0.8includeseyeboltandhydrophonecage.Diameter:16.8.EyeboltAnchor:2.5innerdiameter,4.3outerdiameter,5.1heightoffhousing.StandardHydrophone:6.4length,1.9diameter.

Weight(Dry)withelectronics:9.5withoutbatteries.13.5with32batteries.Buoyancy(saltwater)withelectronics:5.5,withoutbatteries.1.5with32batteries

SoundTrap4channel

n.p. NA NA 70mW9.5x4x3excludinghydrophones

0.12

21x8dia

1SoundTrapHF

n.p. NA NA 35mW9.5x4x2.5excludinghydrohone

21x6dia

WISPR n.p. NA NA Powerusage~800mW8x6.5(digitalunitonly)

<50g(digitalunitonly)

NA NA

VIDEO

CM100 NA NA NA 12W n.p. n.p. n.p. n.p.CM202 NA NA NA na n.p. n.p. n.p. n.p.DualImager NA NA NA n.p. n.p. n.p. n.p. n.p.EO900 NA NA NA n.p. n.p. n.p. n.p. n.p.OTUSU135HIGHDEF

NA NA NA 15W n.p. n.p. n.p. n.p.

OTUS-L205HIGHDEF

NA NA NA n.p. n.p. n.p. n.p. n.p.

TASE310 NA NA NAPower:20W(average)125W(max)

n.p. n.p. n.p. n.p.

203


DATE:July2016




Systemclass

Systemname


Weight(kg)*


TASE400HD NA NA NAPower:40W(average)125W(max)

n.p. n.p. n.p. n.p.

11.3.14 Sensormanningrequirements

Table25.ManningrequirementsforthesensorslistedinTable19.Givenistheminimumnumberofpeopleforoperation,thevesselrequirements,themanningrequirementsfordeploymentandrecoveryaswellasthetrainingneedsforoperatingthesensor.Notethatcellswiththesamekindofinformationforsystemsofthesamemanufacturermaybemerged.Abbreviations:AAM:activeacousticmonitoring,PAM:passiveacousticmonitoring,NA:notapplicable,n.p.:notprovided.

Manningrequirements

Systemclass

SystemnameMinnumberofpeopleforoperation

Vesselrequirements Deployment Recovery Trainingneeds

AAM

Aquadopp NA NA NA NA AcousticprocessingADP NA NA NA NA AcousticprocessingAZFP NA NA NA NA AcousticprocessingDT-XSUB 1 Largeenoughtosupportsmalldavit 2people 2people AcousticprocessingES853 NA NA NA NA AcousticprocessingGemini720i

1 DependsondeploymentframetypeDependsondeploymentframetype

Dependsondeploymentframetype

Lowleveloftraining(1day)Gemini720is

modularVR2C NA NA NA NA Minimal-manualavailableVMT NA NA NA NA

WBAT NA NA NA NAAcousticprocessing

WBTmini NA NA NA NA

PAM

A-Tag NA NA 1person 1person Yes

AUSOMS-miniBlack NA NA 1person 1personNotreally.Itissimplerecordingsystem

C-POD NA NA Canbestartedaheadofmission,then=nil

NilSetupandstart

C-POD-F NA NA MinimalCornell/AutoBuoys NA NA 1 1 n.p.

204


DATE:July2016


Manningrequirements

Systemclass

SystemnameMinnumberofpeopleforoperation

Vesselrequirements Deployment Recovery Trainingneeds

Decimus NA NA Deploymentdependent DeploymentdependentDependsonhowindepththeytogo

DMON NA NA n.p. n.p. n.p.

SDA14 NA NADependsonmetholodogyandcarrier

Dependsonmetholodogyandcarrier

1-daytrainingprovidesautonomytooperator

SDA416 NA NA2-daytrainingprovidesautonomytooperator


NA NA n.p. n.p. n.p.

SM2/SM3 NA NA1-3personsdependingonconditions

1-3personsdependingonconditions

Minimal

SoundTrap4channel NA NA n.p. n.p. n.p.SoundTrapHF NA NA n.p. n.p. n.p.WISPR NA NA 1 1 n.p.

VIDEO

CM100 n.p. NA n.p. n.p. n.p.CM202 n.p. NA n.p. n.p. n.p.DualImager n.p. NA n.p. n.p. n.p.EO900 n.p. NA n.p. n.p. n.p.OTUSU135HIGHDEF n.p. NA n.p. n.p. n.p.OTUS-L205HIGHDEF n.p. NA n.p. n.p. n.p.TASE310 n.p. NA n.p. n.p. n.p.TASE400HD n.p. NA n.p. n.p. n.p.

11.3.15 Sensorfurtherinformation

Table26.FurtherinformationonthesensorslistedinTable19.Givenistherawdatavolume,thecapabilitiestoreducereal-timedata,linksandcomments.Abbreviations:AAM:activeacousticmonitoring,PAM:passiveacousticmonitoring,NA:notapplicable,n.p.:notprovided.

Systemclass

Systemname

RawdatavolumeRealtimedatareductioncapabilities

Links Furthercomments

AAM

Aquadopp32bytes+9*128cellsperping

Boththesontekandthenortekhavebeenputongliders-typicallystrappedonasautonomousinstruments

205


DATE:July2016


Systemclass

Systemname



ADP n.p.

Boththesontekandthenortekhavebeenputongliders-typicallystrappedonasautonomousinstruments

AZFP

TheAZFPcanoperateinaninternallyrecordingmodeoritcanmakedataavailableinrealtime.Dataratesvarysubstantiallyfrom56124to1159bytesperpingdependingonthenumberoffrequencies,binsizeandrange.

n.p.

DT-XSUB 10MB/SecOnboardstoragewithlowbandwidthsummaryreporting

https://www.youtube.com/watch?v=ISMQ5fzm26I

Goodforfittingtowaveglider

https://www.youtube.com/watch?v=3A1aU2CNKhk

https://planetos.com/blog/sonars-are-going-robotic-an-interview-with-biosonics-eric-munday/

https://www.google.com/url?q=http://www.biosonicsinc.com/pdf/BioSonics%2520LR%2520Wave%2520Glider%2520DTX%2520SUB%2520FLIER.pdf&sa=U&ved=0ahUKEwiqkI6xuvDJAhWikKYKHRVlAnYQFggEMAA&client=internal-uds-cse&usg=AFQjCNG2URYhnR2Mt7KLWZvqakPMYUvDOg

ES853 256bytesperping n.p. Guihenetal.2014 Gemini720i

~3600MB/hour

Arollingstoragefacilityisavailable,whereallfilesforthelastweek(forexample)aesaved,butfilesthataredetectedtohaveahighprobabilitytargetaresavedtoapermanentlocation.Also,compressioncanbeappliedtotherecordeddata(ECD)filesthatachieves~50%reductioninsize.

http://www.tritech.co.uk/product/gemini-720i-300m-multibeam-imaging-sonar

Gemini720is

TheGemini720iswillprovideallthecurrent720ifunctionalityandadditionalimprovementsduetotechnologyadvances

modularVR2C

dependsonnumberoftaggedfish-typicallykbyespermonth

n.p.

VMT n.p. WBAT n.p. n.p. TheminiWBTforglider,theWBAT

forotherAUV.Theycanalsobeputonsurfacevehicles.WBTmini n.p. n.p.

206


DATE:July2016


Systemclass

Systemname


Links FurthercommentsPA

M

A-Tag1-128MBdependsonthenumberofdetectedpulseevents

Yes

AUSOMS-miniBlack

2.54GB/hourforLinearPCM yes (dependigonformat)

C-POD 3MB/h yes C-POD-F 30MB/h yes Cornell/AutoBuoys

1GB/hour/channel Yes

Decimus n.p. yes DMON n.p. n.p. SDA14

variable yes

SDA416 Seicherealtimetranmissionsystem

n.p. n.p.

SM2/SM3RecordingMBperhourwill n.p. varybysamplerate.

SoundTrap4channel

upto4GB/h Yes

SoundTrapHF

WISPR880MB/hour/channel/uncompressed

Yes

http://embeddedocean.com/passive-acoustics-2/wispr-v1-0/http://embeddedocean.com/

Alreadyintegratedintoglidersandfloats

VIDEO

CM100 n.p. n.p.http://uavvision.com/product/cm100-html/

CM202 n.p. n.p.http://uavvision.com/product/cm202-3/

DualImager

n.p. n.p.http://www.insitu.com/images/uploads/product-cards/ScanEagle_DualImager_ProductCard_PR041615_1.pdf

EO900 n.p. n.p.http://www.insitu.com/images/uploads/product-cards/ScanEagle_EO900_ProductCard_PR041615_1.pdf

207


DATE:July2016


Systemclass

Systemname



OTUSU135HIGHDEF

n.p. n.p.http://uavvision.com/product/cm100-html/

OTUS-L205HIGHDEF

n.p. n.p.http://www.dst.se/gimbal/otus-l205/otus-l205-high-def

TASE310 n.p. n.p.http://www.cloudcaptech.com/products/detail/tase-310

TASE400HD

n.p. n.p.http://www.cloudcaptech.com/products/detail/tase-400-hd

11.3.16 ListofDataRelaysystems

Table27.DataRelaysystemsthatcanbeusedforautonomousvehicles.

Systemclass Systemname Manufacturer Technologyreadinesslevel

WiFi

900MhzAnalogSeicheBespokeSolution SeicheLtd

Fullcommercialapplication

1800MhzAnalog

3.65GHz NanoStation

UbiqitiWiFi2.4GHzBullet

WiFi5GHz

Satellite

Inmarsat FleetOne CobhamSailor250

Iridium MCG-101 Aurora

Iridium Pilot Iridium

208


DATE:July2016


Systemclass Systemname Manufacturer Technologyreadinesslevel

IridiumL-Band L-Band

209


DATE:July2016


11.3.17 Datarelaytechnicaldetails

Table28.TechnicaldetailsoftheDataRelaysystems:Theiroperationalsizeandweight,power,bandwidth,range,purchaseandoperationalcostsaswellasusagerestrictions,datadelayandfurthercomments.

Systemclass

Systemname

Environmentalperformancelimits

Operationalsize(heightxwidthxdepthincm)

Operationalweight(kg)

Power(W)

Bandwidth

RangePurchasecosts

Operationalcosts

Usagerestriction(geographical,political,etc)

Datadelay

Furthercomments

WiFi

900MhzAnalog Seiche

BespokeSolution

Suitableupto20kmRange

37x30x18 <5 30250kHz

10-20km

n.p. Freespectrum

Geographicrestrictionsapplyonfrequencychoice

Negligible

NoAntennashavebeenincludedastheyareverydependentupondeploymentrequirements

Allsystemsofthisnatureareperformancedependentuponenvironment,antennachoice,antennaheight,lineofsightetc.sowouldrequireevaluatingforsuitabilityonaperprojectbasis.

1800MhzAnalog

n.p.

3.65GHzNanoStation

15km+ 30x30x8 0.2 5.5150Mbps

15+km $150SmallLicencefee

IntheUSyouneedanon-exclusivenationwidelicence,thisismoreaformality.Presentlythebasestationandtransmittermustberegistered.

Negligible

WiFi2.4GHz

Bullet

0-50km,inrealityaround2kmforrequiredbandwidths

15x4x3 0.2 1050Mbps

Upto50km

$120

FreeSpectrum

Operatesonconsumerwififreqsomaybeoverloadedspectrumincertainenvironments

Negligible

WiFi5GHz

Upto50km,lessrangethan2.4Ghzversion

$120

Satellite

Inmarsat FleetOne Global Dia33x28 4 150128kbps

Global $10,000+

From$1,300/Monthfor250MBData

n.p.Negligible

Contractcommitmentmayberequried

210


DATE:July2016


System

classSystemname

Environmentalperformancelimits

Operationalsize(heightxwidthxdepthincm)

Operationalweight(kg)

Power(W)

Bandwidth

RangePurchasecosts

Operationalcosts

Usagerestriction(geographical,political,etc)

Datadelay

Furthercomments

Iridium MCG-101

GlobalReach

Dia18cmx10

1.8 30

ShortBurstDatasolution

GlobalReach

$1,600

$40for30,000bytesofdata(Monthly)

n.p.

Negligible

Iridium Pilot Dia57x20 12.5 100128kbps

Global

$5,000+

From$2,000/monthfor200MBData

n.p.

IridiumL-Band

L-Band n.p.16.2x8.1x2.8

0.422.5(0.2standby)

2400bps

$1099+ n.p.

ContinuousDataConnectionorShortBurstData(SBD)forefficienttransferofsmalldatapacketsupto1960bytes.

small(upto20sforSBD)

211


DATE:July2016


11.4 Evaluationmatrices

11.4.1 UASsensorversusplatformmatrix

Table29.CompilationofwhichVideosensorcouldpotentially(o)ordefinitivelybecombined(x),whichcannotbecombined(-),andwhichcombinationwillworkbutwithlimitedmissionduration(l).NA:notapplicable.

Video

Sensor

type

Sensorname

TASE

400H

D

TASE310

CM20

2

CM10

0

OTU

SU13

5HIGHDEF

OTU

S-L205

HIGHDEF

EO90

0

Dua

lIm

ager

UAS

TrimbleUX5 - - - - - - - -

BRAMORC4EYE

- - - - - - - -

ScanEagle - - - - - - x x

PenguinB - - - o x - - -

Fulmar o o o o o o - -

Jump20 x x o o o o - -

OceanEye - - - - - - - -

SwanX1 NA NA NA NA NA NA NA NA

DesertStar10.

- - - - - - - -

212


DATE:July2016


11.4.2 ASV/AUVsensorversusplatformmatrix

Table30.CompilationofwhichPAMorAAMsensorcouldpotentiallyordefinitivelybecombined,sensor/platformcombinationsthatarealreadyintegrated,whichcombinationscannotbecombinedandwhichcombinationwillworkbutwithlimitedmissionduration.

Key:

PAMsensors AAMsensors

Systemname

A-Tag

AUSO

MS-mini

Black

C-PO

D

C-PO

D-F

Cornell/

AutoB

uoys

Decim

us

DMON

SDA14

SDA41

6

Seiche

realtim

etran

mission

system

SM2/SM

3

Soun

dTrap

4cha

nnel

Soun

dTrapHF

WISPR

Aqu

adop

p

ADP

AZFP

DT-XSU

B

ES85

3

Gem

ini720

i

Gem

ini720

is

mod

ularVR2C

VMT

WBAT

WBTmini

ASV

pow

ered

ASV-6300 o o o o o o o o o o o o o o o o o o o o

C-Cat2 o o o o o o o o o o o o o o o o o o o o o o

C-Enduro o o Y o o Y o o o Y o o o o o o o o o o o o

C-Stat o o o o o o o o o o o o o o o o o o o o

C-Target3 o o o o o o o o o o o o o o o o o o o o o o

C-Worker4 o o o o o o o o o o o o o o o o o o o o o o

Cannotbecombined N NA:Notapplicable

Combinationpossiblewithknownlimitationsinmissiondurationordepth L

Couldpotentiallybecombined o

Coulddefinitelybecombined Y

Combinationsalreadyintegrated Y

Notknownifcombinationispossible

213


DATE:July2016



Systemname

A-Tag

AUSO

MS-mini

Black

C-PO

D

C-PO

D-F

Cornell/

AutoB

uoys

Decim

us

DMON

SDA14

SDA41

6

Seiche

realtim

etran

mission

system

SM2/SM

3

Soun

dTrap

4cha

nnel

Soun

dTrapHF

WISPR

Aqu

adop

p

ADP

AZFP

DT-XSU

B

ES85

3

Gem

ini720

i

Gem

ini720

is

mod

ularVR2C

VMT

WBAT

WBTmini



C-WorkerHydro

o o o o o o o o o o o o o o o o o o o o o o

Delfim o o o o o o o o o o o o o o o o o o o o

Mariner o o o o o o o o o o o o o o o o o o o o

MeasuringDolphin

o o o o o o o o o o o o o o o o o o o o

ROAZI o o o o o o o o o o o o o o o o o o o o

ROAZII o o o o o o o o o o o o o o o L o o o o

RTSYSUSV o o o o o o o o o o o o o o o

ASV

unp

owered

AutoNaut2 o o o o o o o o o o o o o o L L L L L L L




WavegliderSV2

o o o o o Y o o o Y o o Y o L L L L L L L

WavegliderSV3

o o o o o o o o o o o o o o L L L L L L L

AUVglider

ALBAC o o N N N L o L L N N o o o Y Y o o Y Y N Y

Coastalglider o o N N N L o L L N N o o o Y Y o o Y Y N Y

Deepglider o o N N N N o L L N N o o o Y Y o o Y Y N Y

eFòlagaIII o o N N N N o L L N N o o o Y Y o o Y Y N Y

214


DATE:July2016



Systemname

A-Tag

AUSO

MS-mini

Black

C-PO

D

C-PO

D-F

Cornell/

AutoB

uoys

Decim

us

DMON

SDA14

SDA41

6

Seiche

realtim

etran

mission

system

SM2/SM

3

Soun

dTrap

4cha

nnel

Soun

dTrapHF

WISPR

Aqu

adop

p

ADP

AZFP

DT-XSU

B

ES85

3

Gem

ini720

i

Gem

ini720

is

mod

ularVR2C

VMT

WBAT

WBTmini

LiberdadeXwing/Zray

o o N N N N o L L N N o o o Y Y o o Y Y N Y

Petrel o o N N N N o L L N N o o o Y Y o o Y Y N Y

SeaBird o o N N N N o L L N N o o o N N N N N N N N

SeaExplorer o o N N N L o L L N N o o o Y Y o o Y Y N Y

Seaglider(ogive)

o o N N N N Y L L N N o Y Y Y Y o o Y Y N Y

SlocumG2glider

o o N N N L Y L L N N o Y o Y Y o o Y Y N Y

SlocumG2hybrid


SlocumG2thermal


Spray o o N N N N o L L N N o o o Y Y o o Y Y N Y

Sterneglider o o N N N N o L L N N o o o Y Y o o Y Y N Y

TONAI o o N N N N o L L N N o o o Y Y o o Y Y N Y

AUVprope

ller

A18-D o L L o o o L o N o L o o o o o L L L o

A9-M o o L o o o o o N o L o o o o o o o o o

Bluefin-12D o L L o o o o o N o L o o o o o L L L o

Bluefin-12S o o L o o o o o N o o o o o o o o o o o

Bluefin-21 o L L o o o L o N o L o o o o o L L L o

Bluefin-9 o o L o o o o o N o o o o o o o o o o o

Bluefin-9M o o L o o o o o N o o o o o o o o o o o

215


DATE:July2016



Systemname

A-Tag

AUSO

MS-mini

Black

C-PO

D

C-PO

D-F

Cornell/

AutoB

uoys

Decim

us

DMON

SDA14

SDA41

6

Seiche

realtim

etran

mission

system

SM2/SM

3

Soun

dTrap

4cha

nnel

Soun

dTrapHF

WISPR

Aqu

adop

p

ADP

AZFP

DT-XSU

B

ES85

3

Gem

ini720

i

Gem

ini720

is

mod

ularVR2C

VMT

WBAT

WBTmini


o o L o o o o o o N o o L o o o o o o L o o


o L L o o o L o o N o o L o o o o o L L L o

HUGIN3000 o L L o o o L o o N o o L o o o o o L L L o

HUGIN4500 o L L o o o L o o N o o L o o o o o L L L o

MUNIN o L L o o o o o o N o o L o o o o o L L L o

REMUS100 o o o o o o o o o N o o o o o o o o o o o o

REMUS3000 o L L o o o o o o N o o L o o o o o L L L o

REMUS600 o o L o o o o o o N o o L o o o o o o L o o

REMUS6000 o L L o o o L o o N o o L o o o o o L L L o

RTSYSAUV o o L o o o o o o N o o o o o o o o o

216


DATE:July2016


11.5 Suppliercontacts

Table31.Listofsuppliercontactsoftheplatformsandsensorsincludedinthisreview.

Systemclass

Manufacturer Systemname(s) Street+No Town Postcode Country Phonenumber website

ASV

-po

wered

ASVC-Enduro,C-Worker4+6+7+Hydro,C-Cat2,C-Target3,C-Stat,ASV-6300

Unit12MurrillsEstate,SouthamptonRoad

Portchester PO169RD UK T+44(0)2392382573 asvglobal.com

InstitutoSuperiordeEngenhariadoPorto

ROAZIandIIRuaDr.AntónioBernardinodeAlmeida

Porto431,4200-072

Portugal T+351228340500 www.isep.ipp.pt

RTSYS SeawaysUSV RueMichelMariono 56850Caudan France T+33297898582 www.seaways.fr

RostockUniversity MeasuringDolphinAnwendungszentrumRegelungstechnik,Tannenweg22d,imRostockpark

Rostock 18059 Germany T+493814987727 www.uni-rostock.de

InstitutoSuperiorTécnico

DelfimTorreNorte,7thFloor.Av.RoviscoPais

Lisbon1.1049-001

Portugal T+351218418289 tecnico.ulisboa.pt

MaritimeRoboticsAS Mariner Brattørkaia11 Trondheim 7010 Norway T+4773401900 www.maritimerobotics.com

ASV

-un

powered

LiquidRobotics WavegliderSV2+3 1329MoffettParkDr Sunnyvale94089-1134

UnitedStates

T+14086364200 liquidr.com

MOST(AutonomousVessels)Ltd

AutoNaut2+3+5+7UnitA5TheBoatyard,ChichesterMarina

Chichester PO207EJ UK T+44(0)1243511421 www.autonautusv.com

AUV-glider

ALSEAMAR–ALCEN Seaexplorer 9Europarc 13590Meyreuil France T+33442484452 www.acsa-alcen.com

EcoleNationaleSuperioreD'IngenieursBrest

Sterneglider Nocontactdetails NocontactdetailsNocontactdetails

Nocontactdetails

Nocontactdetails Nocontactdetails

ExocetusDevelopmentLLC

Coastalglider 1444East9thAve Anchorage 99501 USA 9072278073 exocetus.com

Graaltech Folaga viJ.Ruffini9/r 16128Genova Italy T+390108597680 www.graaltech.com

KongsbergSeaglider(normal-alsoenquiredeep)

1921033rdAvernueWest SeattleWA98036-4707

USA T+14257121107 www.km.kongsberg.com

KyushuInstituteofTechnology

SeaBird 1-1SensuichoTobataWard,Kitakyushu

804-0015 Japan T+81938843000 www.kyutech.ac.jp

Nortek Aquadopp Vangkroken2 1351Rud Norway T+4767174500 www.nortek-as.com

217


DATE:July2016


Systemclass


SCRIPPSSprayGlider,LiberdadeXwing/Zray

9500GilmanDrive LaJolla92093-0230

USA (858)4614728 www.scripps.edu

TeledyneWebbSlocumglider(electric,thermalandhybrid)

49EdgertonDrive NorthFalmouth MA02556 USA T+15085482077 www.teledyne.com

TianjinUniversity Petrel Nocontactdetails NocontactdetailsNocontactdetails

China Nocontactdetails Nocontactdetails

TokaiUniversity ALBAC 2-28-4Tomigaya Shibuya 151-0063 Japan T+81334672211 www.u-tokai.ac.jp

OsakaPrefectureUniversity,TaijiWhaleMuseum,Cetus

TONAI:TwilightOcean-ZonalNaturalResourcesandAnimalInvestigator

Nocontactdetails NocontactdetailsNocontactdetails

Japan Nocontactdetails Nocontactdetails

AUV-prop

eller BluefinRobotics Bluefin9,9M,12D,12S,21 553SouthStreet Quincy 2169 USA T+1(617)7157000 www.bluefinrobotics.com

ECAGroup A9-M,A18-D Z.I.ToulonEst262,ruedesfrèresLumière

83130LaGarde

France T+33(0)494089000 www.ecagroup.com

KonsbergMaritime REMUS,HUGIN,MUNIN Kirkegårdsveien45 NO-3616Kongsberg NorwayT+4732285000

www.km.kongsberg.com

RTSYS AUVrobots RueMichelMariono 56850Caudan France T+33297898582 www.rtsys.eu

UASkite

FlyingRobotsSA SwanX1 RueThalberg,2 Geneva CH-1201Switzerland

www.flying-robots.com

UAS

l-t-a AllsoppHelikitesLtd DesertStar

Unit2,FordingbrigdeBusinessPark

Fordingbridge,Hampshire

SP61BD UK T+441425654967 www.allsopp.co.uk

MaritimeRoboticsAS OceanEye Brattørkaia11 Trondheim 7010 Norway T+4773401900 www.maritimerobotics.com

UAS-p

owered

-fixed

ArcturusUAV Jump20 P.O.Box3011 RohnertPark 94928 USA (707)2069372 arcturus-uav.com

C-Astrald.o.o. BRAMORC4EYE,gEO,rTK Gregorčičevaulica20 Ajdovščina 5270 Slovenia T+386(0)40121119 www.c-astral.com

Insitu ScanEagle 118EastColumbiaRiverWay Bingen,Washington 98605 USA T+15094938600 www.insitu.com

ThalesEspañagrp,S.A.U.

Fulmar SerranoGalvache,56 Madrid 28033 Spain T+34912737200 www.thalesgroup.com

TrimbeNavigationLimited

UX5 Buchtenstraat9/1 Gent 9051 Belgium T+3293350515 uas.trimble.com

218


DATE:July2016


Systemclass


Uavfactory PenguinB 50SouthBuckhoutStreet Irvington 10533 USA T+1(914)5913296 www.uavfactory.com

AAM

ASL AZFP #1-6703RajpurPlace Victoria,BC V8M1ZS Canada T+1-250-656-0177 www.aslenv.com

BiosonicsAutonomoussubmersibleechosounder

4027LearyWayNW Seattle 98107 U.S.A. 2069636369 biosonicsinc.com

Imagenex ES853 209-1875BroadwayStreet PortCoquitlam, V3C4Z1 Canada 604-944-8248 www.imagenex.com

Kongsberg/Simrad WBAT+WBTmini P.O.Box111Strandpromenaden50

3191 Norway 4733034026 www.simrad.com

Nortek Aquadopp Vangkroken2 1351Rud Norway T+4767174500 www.nortek-as.com

Sontek ADP 9940SummersRidgeRoad SanDiego CA USA T+1(858)5468327 www.sontek.com

Tritech Gemini720i&720isPeregrineRoad,WesthillBusinessPark

Westhill AB326JL UK T+44(0)1224744111 www.tritech.co.uk

Vemco modularVR2C+VMT 20AngusMortonDrive Bedford,NovaScotia B4B0L9 Canada T+1-9024501700 vemco.com

PAM

AquasoundInc. AUSOMS-miniBlackKyotoResearchCenter,46Shimoadachi-cho

YoshidaSakyo-ku 606-8501 Japan T+81-0757539670 aqua-sound.com

Chelonia CPOD+C-POD-F TheBarkhouse,NorthCliff Mousehole TR196PH UK T+44(0)1736732462 www.chelonia.co.uk

Cornell(PAM)/EOS(buoy)

Cornell/AutoBuoys/DOSITSCornellLabofOrnithology,159SapsuckerWoodsRd.

Ithaca 14850 USA 607-2546250 www.birds.cornell.edu

EmbeddedOceanSystems

WISPR Seattle USA (206)7660671 embeddedocean.com

JASCO AMARTheRoundel,StClair’sFarm,WickhamRoad,

Droxford SO323PW UK (0)1489878439 www.jasco.com

MarineMicroTechnology

A-Tag 4-12-1Takakura IrumaCity Japan 0429654127

OceanInstrumentsNZ SoundTrapHF+4channel 961SandspitRd Warkworth 982NewZealand

6499233601 www.OceanInstruments.co.nz

RTSYS PAM:RB-SDA14,BA-SDA14 RueMichelMariono 56850Caudan France 33297898582 www.rtsys.eu

SAInstrumentationLtd Decimus MillCourt,MillLane Tayport DD69EL UK T+44(0)1334845260 www.sa-instrumentation.com

SeicheMeasurementsLtd


BradworthyIndustrialEstate,LangdonRoad,Bradworthy

Holsworthy EX227SF UK T+44(0)1409404050 www.seiche.com

WHOI DMON 266WoodsHoleRoad WoodsHole02543-1050

USA T+15082892906 www.whoi.edu

219


DATE:July2016


Systemclass


WildlifeAcoustics SM2/SM3 3ClockTowerPlace,Suite210 Maynard01754-2549

USA T+1(978)3695225 www.wildlifeacoustics.com

Video

CloudcapTechnology TASE400HD+310 202WascoLoop,Suite103 HoodRiver 97031 USA (541)3872120 www.cloudcaptech.com

DSTOTUSU135+L205HIGHDEF

Åkerbogatan10 Linköping 582 Sweden T+4613211080 www.dst.com

Insitu EO900+DualImager 118EastColumbiaRiverWay Bingen,Washington 98605 USA T+15094938600 www.insitu.com

UAVvision CM202+100 10UrallaStreet PortMacquarie 2444 Australia T+61265811994 uavvision.com

220


DATE:July2016


11.6 Surveyingwildlifepopulations

Thissectionprovidesfurtherdetailsonthestatisticalsideofdensityestimatesexplaininghowsurveydataare

analysed (section11.6.1)andwhichmethodsarecurrentlyavailable forestimating thedetectionprobability

(section11.6.2),whichservesforabetterunderstandingontherelevanceofretrievingappropriatesurveydata

fromacarefullyplannedandexecutedsurveydesign.

11.6.1 Estimatingabsolutepopulationdensityandabundance–anoverview

Statistical data analysismethods have been developed that correct observed data for undetected animals,

leadingtoabsoluteestimatesofanimalabundanceordensity(Borchersetal.,2002).Atthedataanalysisstage,

estimatorsareusedtoproducetherequireddensityorabundanceestimates.Atypicalestimatorforabsolute

densityis:

! = #$% (Eqn.1)

Where

!istheestimateddensity,

nisthenumberofencountersmadeduringthesurvey,

&istheestimatedaverageprobabilityofdetectinganencounterand

aisthetotalsurveyedarea.

It is rare that an entire study areaof interest canbe completely coveredby a survey (due to logistical and

financialconstraints)soitisoftenthecasethatthesurveyedareaisasubsampleofthestudyarea.Absolute

abundance in the study area can be estimatedusing Eqn. 2, if the survey has beendesigned such that the

surveyedareaisrepresentativeofthestudyarea,otherwiseresultingabundanceestimatesmaybebiased.

' = !×) (Eqn.2)

Where

'istheabsoluteabundanceinthestudyareaand

Aisthesizeofthestudyarea.

Theabilitytorelyonasurvey’sdesigntoinferthatestimatedanimaldensityisrepresentativeofdensityinthe

widerstudyarea,andthatabundanceinthestudyareacanbeestimatedusingEqn.2,isknownasdesign-based

inference (Borchers et al., 2002). An adequate survey design comprises multiple samplers such as survey

transects,whichhavebeenrandomlysampledfromallpossiblesamplersinthestudyarea(orstratum,ifthe

surveyisstratifiedgeographically)withequalprobability.Asystematicrandomdesignistypicallypreferableto

acompletelyrandomdesignbecauseitproducesmorepreciseestimates.Furthermore,thesamplersshouldbe

221


DATE:July2016


randomlydistributedwithrespecttothestudyspecies’distribution(seeBucklandetal.,2001;2015formore

detailaboutsurveydesign). Ifasurveyhasnotbeendesigned,or therealisedsurveyhasdeviated fromthe

planneddesign, thenmodel-based inferencecanbeused insteadto inferdensityandabundanceacross the

studyarea.Thisapproachusesastatisticalmodelto linkestimateddensity inthesurveyedareawithspatial

covariates.Themodelcanthenbeusedtopredictthedensity/abundanceinunsurveyedpartsofthestudyarea

(e.g.CañadasandHammond,2006;Miller,2013).However,whilethisapproachrelaxestheneedforastrict

surveydesign,thereisstillarequirementthat,foranyselectedcovariate,therangeofmodelledvaluesmatches

therangeofvaluesinthestudyarea.Biasedresultsmayoccurifthespatialmodelisusedtogenerateabundance

predictionsusingcovariatevaluesthatarelowerorhigherthanthevaluesusedtobuildthemodel(i.e.model

extrapolation).Therefore,thesurveyedareamustberepresentativeofthewiderstudyarea.

Thedefinitionofanencountercanchangedependingonthespeciesofinterestandthesurveytype.However,

density estimators are flexible and can be altered accordingly. In some cases, it may be possible to count

individuals(visuallyoracoustically)butinothersurveysofhighlysocialorshoalinganimals,groupsofanimals

may form the observed encounters. When groups are counted, average group size can be included as an

additionalmultiplyingparameter,or“multiplier”,inthedensityestimatortoconvertestimatedgroupdensity

into estimated animal density. In the caseof shoaling fish and zooplankton,where group size estimation is

particularlydifficult,thereisalargeliteratureontheconversionofechosounderdatatoestimatesofdensityor

abundance(e.g.JohannessonandMitson,1983;Foote,2009).Alternatively,averagebiomassmaybeestimated

(e.g.Coxetal.,2011). Inpassiveacousticmonitoring,callscanoftenbereadilycountedandanaveragecall

productionrateisthenrequiredtoconvertestimatedcalldensitytoestimatedanimaldensity.Furthermore,

acousticdataareoftenautomaticallyprocessedandanadditionalmultiplierisrequiredtocorrectthenumber

ofobservationsforfalsedetections.Atypicaldensityestimatorforusewithpassiveacousticdatais,therefore:

! = # *+,$%- (Eqn.3)

Where

nisthenumberofacousticencounters,

fistheproportionoffalsedetectionsgeneratedbythedetectionroutineand

.representstheappropriatemultiplier(s)thatwillconverttheobjectdensitytoanimaldensity.

If, forexample, individualcallswerecounted inasurvey,thenrequiredmultiplierswouldbeanaveragecall

productionrate(anestimatedparameter)andtheamountoftimespentmonitoring(aconstant).

Theprobabilityofdetection typically correctsencounters forperceptionbias. This isbias causedbymissing

encountersthatwereavailablefordetectionbutwerenotdetectede.g.becauseofhighseastateduringvisual

surveysorhighbackgroundnoiseinacousticsurveys.Anotherformofbiasthatmustbeconsideredisavailability

bias.Thisisbiasthatarisesfrommissinganimalsthatwerepresentinthesurveyareabutwerenotavailableto

222


DATE:July2016


bedetected.Invisualsurveysfromsurfacevesselsoraerialplatforms,animalsareonlyavailabletobeseenat

thesurfaceoratveryshallowdepthsandarethereforeonlytemporarilyavailableduringthesurvey.Therefore,

anadditionalmultiplierisrequiredinthedensityestimatortoaccountfortheproportionofanimalsthatwere

unavailabletobesurveyed.Similarly,inacousticsurveys,animalsarenotconstantlyvocal,soamultiplier,such

ascallproductionrate,isrequiredtoaccountforanimalsinthepopulationthatwerenotvocalisingatthetime

ofthesurvey.Furthermore,someanimalswillbepermanentlyunavailableinpassiveacousticsurveysifthey

neverproducethespecificvocalisation(s)thatarebeingmonitored.Thisproportionofthepopulationmustalso

beaccountedforinthedensityestimatorusinganappropriatemultiplier.Itisimportanttorecognisethatany

estimatedmultiplier (including theprobabilityofdetection)may showspatial and temporal variation.Using

multipliers from the literature or that have been estimated from datasets collected from another study

population (or even the same study population but at a different location and/or time) may result in the

applicationofthewrongmultiplierand,ultimately,biasedresults.

11.6.2 Methodstoestimatedetectionprobability

Wildlifesurveystypicallytakeplacealongplannedsurveytransectsorfromstaticmonitoringpoints.Insome

surveys,itmaybepossibletodetectallanimalswithinthetransectlinesorpoints.Suchsurveytypesareknown

asstriptransectsampling(usingtransectlines)orplotsampling(usingpoints).

Whenanimalsaresuspectedtobemissedwithinsurveyedareas,thereareseveralmethodsthatcanbeusedto

estimatetheprobabilityofdetection.Distancesampling,forexample, isacommonly-used,versatilemethod

thathasbeenwidelyappliedtomarinesurveys(Bucklandetal.,2001;2015).Distancesamplinghasbeenapplied

tovisualandacoustic(bothactiveandpassive)datatoestimatetheabundanceofmarinemammals,turtlesand

krillswarms.Distancesamplingdatahavebeencollectedfromavarietyofmarinesurveyingplatformsincluding

vessels and aircraft that follow survey lines (line transect sampling) and stationarymonitoring points (point

transectsampling).Detectionprobabilitiesestimatedusingdistancesamplingcanbeusedindensityestimators

suchasthoseinEqns1and3.

Mark-recaptureisanotherstandardabundanceestimationmethod(Borchersetal.,2002)buthasanalternative

estimation framework to distance sampling, involving different estimators to those in Eqns. 1 & 3. Mark

recapturehasbeenusedwithvisualdata toestimatemarinemammal (e.g.Cheneyetal.,2014), turtle (e.g.

Duttonetal.,2005)andfishabundances(e.g.Bradshawetal.,2007).Datafromanimal-borneactiveacoustic

tagscanalsobeanalysedusingofmark-recapture;taggedanimalsarere-identifiedastheyswimpastdeployed

acousticreceivers(e.g.Dudgeonetal.,2015).Adifficultyofmark-recapturemethodsisthat,whileabundance

canbeestimated,it isnon-trivialtodeterminethesizeofthesurveyedareaand,therefore,estimateanimal

density.However,spatially-explicitcapture-recapture(SECR),arelativelyrecentextensiontomark-recapture

methodology, allowsdensity, aswell as abundance, tobe inferred from the collecteddata.Despitebeinga

223


DATE:July2016


recentlydevelopedmethod,SECRisconsideredtobea“standard”abundanceanddensityestimationapproach

duetoitslinkswithmark-recaptureandalsodistancesampling(Borchers,2012;Borchersetal.,2015).

SECRhasbeenusedwithbothvisual(Pirottaetal.,2014)andpassiveacoustic(Marquesetal.,2011;Martinet

al.,2012)datatoestimatetheabundanceanddensityofmarinemammals.Theabilitytoestimatedensityisa

majoradvantageofSECRovermark—recaptureand,therefore,SECRwillbediscussedinpreferencetomark-

recaptureintherestofthereport.

IfdistancesamplingorSECRcannotbeused,detectionprobabilitiescanbeestimatedusingalternative“non-

standard” methods. Many of the non-standard methods were developed specifically for use with passive

acousticdatatoestimatemarinemammaldensity.Thesemethodsoftenuseauxiliaryinformation.Forexample,

a“trial-based”methodimplementedbyMarquesetal.,(2009)usedpassiveacoustictagdata(specificallyDTag

data:Johnson&Tyack.,2003)tocrossreferencewhetherclicksproducedbythetaggedanimalwerereceived

ornotonfixedhydrophonesinthearea.Theprobabilityofdetectingabeakedwhaleonafixedhydrophone

couldthenbemodelledasafunctionofrange(Marquesetal.,2009).

The various methods require different information to be recorded about encounters. In general, distance

samplingrequiresthatthehorizontalandperpendicularrangeismeasuredfromthetransectlineorpointto

eachencounter.Rangeestimationmethodsusingpassiveacousticdatathatcontainambiguityaboutwhether

theencounterwasontheleft-orright-handsideofthetransectline,orthebearingfromapoint,areacceptable.

However,anestimateofhorizontalrangerequiresthedepthofacoustically-detectedanimalstobeestimated

(but also see Cox et al., 2011 and Harris, 2014 for alternative approaches to extending standard distance

samplingtoaccountforanimalsatdepth).

Spatially-explicitcapture-recapturewhenusedwithvisualdatarequiresthat individualsarere-identified, i.e.

“re-captured” across surveying occasions and that the location of each encountered animal is recorded. In

passiveacousticsurveys,thedatarequirementsforSECRaresomewhatdifferent.Hydrophonesareconsidered

tobe“traps”ofknownlocation,andthesameacousticencountercanbe“caught”onmultiplehydrophones.

The spatial pattern of captures of a given acoustic encounter provides information about the detection

probabilityoftheencounter.Receivedlevelandtime-of-arrivaloftheacousticencounterscanalsoimprovethe

precisionoftheresults(Stevensonetal.,2015).

Non-standardmethodscanbeimplementedwhenthereislessinformationavailableabouttheencounter(i.e.,

ranges,bearingsorrecapturescannotberecorded)butthemethodsrelyonmoreassumptionsandmodelling

tocompensateforthelackofempiricaldataandshouldnotbeconsideredinpreferencetostandardapproaches.

Depending on the particular method being used, measures of received levels and noise levels, as well as

estimates of transmission loss (estimated using sound propagation models) may be useful. These acoustic

measurescanbecombinedinasimulationframeworkthatcanbeusedtoestimatedetectionprobability(e.g.

Kuseletal.,2011).Finally,likestandardmethods,anyadditionalmultipliersthathavebeenidentifiedasrelevant

foragivensurvey,e.g.callproductionrateorgroupsize,willalsoberequired.

224


DATE:July2016


Eachdensityestimationmethodhasassociatedassumptionsofvaryingseverity.Violationofkeyassumptions

canleadtobiasinestimateddetectionprobabilitiesand,ultimately,abundanceordensityestimates.Thereare

fourmainassumptionsindistancesampling(Bucklandetal.,2015).Theseare:(1)thattransectlinesorpoints

havebeenplacedrandomlywithrespecttothedistributionofthestudyspecies(thisisimportantnotonlyfor

thedesign-basedabundanceinference,butalsoforthedetectionprobabilityestimation),(2)anyencounteron

(i.e.,atzerodistancefrom)thetransectlineoratthecentreofthemonitoringpointisdetectedwithcertainty,

(3)thatencounteredanimalsaredetectedattheirinitiallocationi.e.,thereisnoanimalmovementand(4)that

measureddistancesareaccurate.ThemainassumptionofSECRarethatcaptures(whetherindividualanimals

oracousticencounters)canbeaccuratelyre-identifiedacrosscaptureoccasions.WhenusingSECRwithpassive

acousticdata,itiscurrentlyassumedthat(1)allcallsareemittedatthesamesourceleveland(2)thatthereis

noeffectofsignaldirectionalityondetectionprobability(Stevensonetal.,2015).Non-standardmethodsvary

intheirassumptionsbutasharedimportantassumptionisthatanyauxiliarydatausedintheanalysisisrelevant

tothestudypopulationatthetimeandplaceofthestudy.

The final aspect of abundance and density methodology to discuss is variance estimation. Estimating the

uncertainty in a density or abundance estimate is a crucial part of an analysis. It is important to generate

estimatesthatareaspreciseaspossiblesothattheycanbeusefulinmanagementandmitigationdecisions.The

deltamethodisaconvenientwayinwhichtheuncertaintyassociatedwitheachparameterinagivendensity

estimatorcanbecombinedtoestimateanoverallvarianceforthedensityorabundanceestimate(Seber,1982).

225


DATE:July2016


12 Glossary of Terms, Acronyms and Abbreviations Term Description

AAM ActiveAcousticMonitoring

ADCP AcousticDopplerCurrentProfiler

AMG AircraftManagementGuidelines

AMOC AlternativeMethodOfCompliance

ASV AutonomousSurfaceVehicles

ATC AirTrafficControl

AUV AutonomousUnderwaterVehicles

BAS BritishAntarcticSurvey

BLOS BeyondVisibleLineOfSight

bps,kbps,Mbps Bitspersecond,Kilobitspersecond,Megabitspersecond.N.B.8bitsarerequiredto

formoneByteofdata.Mostcommunicationssystemsquotebandwidthinbps.

COTS CommercialOff-The-Shelf

CREEM CentreforResearchintoEcologicalandEnvironmentalModelling

CT Condutivity-Temperatureprobe

DS DepthSensor

DTAG DigitalAcousticRecordingTag

DVL DopplerVelocityLogger

E&P ExplorationAndProduction

ERMA EnergyRatioMappingAlgorithm

EVLOS ExtendVisibleLineOfSight

FAA FederalAviationAdministration

FLAC FreeLosslessAudioCodec

FLS Forward-LookingSonar

GENIoS Glider-EnvironmentNetworkInformationSystem

226


DATE:July2016


Term Description

GPS GlobalPositioningSystem(simpleordifferential)

HSE HealthAndSafetyExecutive

IMU InertialMeasurementUnit

INS InertialNavigationSystem

IR Infrared

Iridium Iridiumsatellitecommunications

ISAS InterferometricSyntheticApertureSonar

LBL LongBaseLine

LFDCS LowFrequencyDetectionAndClassificationSystem

LIDAR LightImaging,Detection,AndRanging

MBES MultiBeamEchoSounder

MMO MarineMammalObserver

NCAA NationalCivilAviationAuthority

NORUT NorthernResearchInstitute

PAM PassiveAcousticMonitoring

PAR PhotosyntheticallyActiveRadiation

PSO ProtectedSpeciesObserver

RF RadioFrequencylink

RHIB Rigid-HulledInflatableBoat

RiNKO TypeofphosphorescentDOsensor

RPAS RemotelyPilotedAircraftSystems

SAS SyntheticApertureSonar

SBL ShortBaseLine

SBP Sub-BottomProfiler

SMRU SeaMammalResearchUnit

227


DATE:July2016


Term Description

SSS SideScanSonar(singleordualfrequency)

SUNA SubmersibleUltravioletNitrateAnalyzer

SVTP Sound,Velocity,TemperatureandPressure

SVS SoundVelocitySensor

TS TurbiditySensor

UAS UnmannedAerialSystems

UNCLOS UnitedNationsConventionOnTheLawOfTheSea

USBL Ultra-ShortBaseLine

UTP UnderwaterTransponderPositioning

VLOS VisibleLineOfSight

WHOI WoodsHoleOceanographicInstitution

WiFi WiFilink

228


DATE:July2016


13 Reference list Acevedo-Whitehouse, K., A. Rocha-Gosselin, and D. Gendron. 2010. A novel non-invasive tool for disease

surveillanceoffree-rangingwhalesanditsrelevancetoconservationprograms.AnimalConservation13:217-

225.

ACSA.SeaExplorerMulti-MissionGlider.2013.

Ainley,D.G.,G.Ballard,R.M. Jones,D. Jongsomjit,S.D.Pierce,W.O.Smith Jr,andS.Veloz.2015.Trophic

cascadesinthewesternRossSea,Antarctica:revisited.MarEcolProgSer534:1-16.

Altan,M.O., T.M. Celikoyan, G. Kemper, andG. Toz. 2004. Balloon photogrammetry for cultural heritage.

InternationalArchivesofthePhotogrammetry,RemoteSensingandSpatialInformationSciences35:964-968.

Alvarez, A., A. Caffaz, A. Caiti, G. Casalino, L. Gualdesi, A. Turetta, and R. Viviani. 2009. Folaga: A low-cost

autonomousunderwatervehiclecombininggliderandAUVcapabilities.OceanEngineering36:24-38.

Alves,J.,P.Oliveira,R.Oliveira,A.Pascoal,M.Rufino,L.Sebastiao,andC.Silvestre.VehicleandMissionControl

oftheDELFIMAutonomousSurfaceCraft.InstituteforSystemsandRobotics.2002.

Antunes,R.,P.H.Kvadsheim,F.P.A.Lam,P.L.Tyack,L.Thomas,P.J.Wensveen,andP.J.O.Miller.2014.High

thresholds for avoidance of sonar by free-ranging long-finned pilot whales (Globicephala melas). Marine

PollutionBulletin.

Araki,S.andK.Ishii.DevelopmentofglidertypesmallAUVSeaBird.UnderwaterTechnologyandWorkshopon

ScientificUseofSubmarineCablesandRelatedTechnologies,2007.Symposiumon,320-325.2007.IEEE.

Armstrong,R.A.,H.Singh,J.Torres,R.S.Nemeth,A.Can,C.Roman,R.Eustice,L.Riggs,andG.Garcia-Moliner.

2006.CharacterizingthedeepinsularshelfcoralreefhabitatoftheHindBankmarineconservationdistrict(US

VirginIslands)usingtheSeabedautonomousunderwatervehicle.ContinentalShelfResearch26:194-205.

Asakawa,K.,M.Nakamura,T.Kobayashi,Y.Watanabe,T.Hyakudome,Y.Ito,andJ.Kojima.Designconceptof

Tsukuyomi-Underwatergliderprototypeforvirtualmooring.OCEANS,2011IEEE-Spain,1-5.2011.IEEE.

Au,W.W.1996.Acousticreflectivityofadolphin.TheJournaloftheAcousticalSocietyofAmerica99:3844-

3848.

Au,W.W.andK.Banks.1998.TheacousticsofthesnappingshrimpSynalpheusparneomerisinKaneoheBay.

TheJournaloftheAcousticalSocietyofAmerica103:41-47.

Baldacci, A., M. Carron, and N. Portunato. 2005. Infrared detection of marine mammals. NATO Undersea

ResearchCentreTechnicalReportSR-443.NATOUnderseaResearchCentre,NewYork,NewYork,USA.

Bauer, R. K., S. Bonhommeau,B. Brisset, and J.M. Fromentin. 2015.Aerial surveys tomonitor bluefin tuna

abundanceandtrackefficiencyofmanagementmeasures.MarEcolProgSer534:221-234.

229


DATE:July2016


Baumgartner,M.F.2014.Real-timepassiveacousticmonitoringofbaleenwhalesfromautonomousplatforms.

TheJournaloftheAcousticalSocietyofAmerica136:2116-2117.

Baumgartner,M.F.andD.M.Fratantoni.2008.Dielperiodicity inbothseiwhalevocalization ratesand the

verticalmigrationoftheircopepodpreyobservedfromoceangliders.LimnologyandOceanography53:2197-

2209.

Baumgartner,M.F.,D.M.Fratantoni,T.P.Hurst,M.W.Brown,T.V.Cole,S.M.VanParijs,andM.Johnson.

2013. Real-time reporting of baleen whale passive acoustic detections from ocean gliders. Journal of the

AcousticalSocietyofAmerica134:1814-1823.

Baumgartner,M.F.,K.M.Stafford,P.Winsor,H.Statscewich,andD.M.Fratantoni.2014b.Glider-basedpassive

acousticmonitoringintheArctic.MarineTechnologySocietyJournal48:40-51.

Baumgartner,M.F., S.M.VanParijs, F.W.Wenzel,C. J.Tremblay,H.C.Esch,andA.M.Warde.2008. Low

frequencyvocalizationsattributedtoseiwhales(Balaenopteraborealis).124:1339-1349.

Baumgartner,M.,T.Hurst,J.Partan,andL.Freitag.TheDMON2:ACommerciallyAvailableBroadbandAcoustic

MonitoringInstrument.2014a.DTICDocument.

Bender,A.,D.M.Steinberg,A.L.Friedman,andS.B.Williams.Analysisofanautonomousunderwaterglider.

ProceedingsoftheAustralasianConferenceonRoboticsandAutomation,1-10.2008.

Benoit-Bird, K. and W. Au. 2003a. Hawaiian spinner dolphins aggregate midwater food resources through

cooperativeforaging.TheJournaloftheAcousticalSocietyofAmerica114:2300.

Benoit-Bird,K.J.andW.W.Au.2003b.Preydynamicsaffectforagingbyapelagicpredator(Stenellalongirostris)

overarangeofspatialandtemporalscales.BehavioralEcologyandSociobiology53:364-373.

Benson,S.R.,K.A.Forney, J.T.Harvey, J.V.Carretta,andP.H.Dutton.2007.Abundance,distribution,and

habitatofleatherbackturtles(Dermochelyscoriacea)offCalifornia,1990-2003.FisheryBulletin105:337-347.

Bernasconi,M.,L.Nøttestad,B.R.E.Axelsen,andJ.O.Krakstad.2011.Acousticobservationsofduskydolphins

Lagenorhynchus obscurus hunting Cape horsemackerel Trachurus capensis off Namibia.Mar Ecol Prog Ser

429:209-218.

Bernasconi,M.,R.Patel, L.Nottestad,G.Pedersen,andA.Brierley.2013.Theeffectofdepthon the target

strength of a humpback whale (Megaptera novaeangliae). Journal of the Acoustical Society of America

134:4316-4322.

Bevan,E.,T.Wibbels,B.M.Najera,M.A.Martinez,L.A.Martinez,F.I.Martinez,J.M.Cuevas,T.Anderson,A.

Bonka,andM.H.Hernandez.2015.Unmannedaerialvehicles(UAVs)formonitoringseaturtlesinnear-shore

waters.MarineTurtleNewsletter145:19-22.

230


DATE:July2016


Bingham,B.,N.Kraus,B.Howe,L.Freitag,K.Ball,P.Koski,andE.Gallimore.2012.Passiveandactiveacoustics

usinganautonomouswaveglider.JournalofFieldRobotics29:911-923.

Blidberg,D.R.ThedevelopmentofAutonomousUnderwaterVehicles (AUV);Abrief summary. IEEE ICRA4.

2001.

Block,B.A.,S. L.Teo,A.Walli,A.Boustany,M. J. Stokesbury,C. J. Farwell,K.C.Weng,H.Dewar,andT.D.

Williams.2005.ElectronictaggingandpopulationstructureofAtlanticbluefintuna.Nature434:1121-1127.

Bonggay,C.CommercialDroneRulesAroundtheWorld.PrecisionHawkInc.2016.2-10-2015.

Borchers, D. 2012. A non-technical overview of spatially explicit capture-recapture models. Journal of

Ornithology152:435-444.

Borchers,D.andD.Kidney.Flexibledensitysurfaceestimationforspatiallyexplicitcapture-recapturesurveys.

CREEMTechnicalReport,no.2014-1.2014.

Borchers,D.L.,B.C.Stevenson,D.Kidney,L.Thomas,andT.A.Marques.2015.Aunifyingmodelforcapture-

ecaptureanddistancesamplingsurveysofwildlifepopulations.JournaloftheAmericanStatisticalAssociation

110:195-204.

Borchers,D.L.,S.T.Buckland,andW.Zucchini.2002.Estimatinganimalabundance:closedpopulations.13ed.

SpringerScience&BusinessMedia.

Bostrom, BL. Jones, TT. Hastings, M. Jones, DR. 2010. Behaviour and Physiology: The Thermal Strategy of

LeatherbackTurtles.PLoSONE5(11):e13925.doi:10.1371/journal.pone.0013925.

Bourgeois,B.,M.Harris,P.Alleman,J.Cheramie,andJ.Gravley.AutonomousBathymetricSurveying.1997.DTIC

Document.

Bradshaw,C.J.,H.F.Mollet,andM.G.Meekan.2007.Inferringpopulationtrendsfortheworld'slargestfish

frommark-recaptureestimatesofsurvival.JournalofAnimalEcology76:480-489.

Bresette,M.J.,B.E.Witherington,R.M.Herren,D.A.Bagley,J.C.Gorham,S.L.Traxler,C.K.Crady,andR.

Hardy.2010.Size-classpartitioningandherdinginaforaginggroupofgreenturtlesCheloniamydas.Endangered

SpeciesResearch9:105-116.

Brierley, A. S. and P. G. Fernandes. 2001. Diving depths of northern gannets: acoustic observations of Sula

bassanafromanautonomousunderwatervehicle.TheAuk118:529-534.

Brierley,A.S.,P.G.Fernandes,M.A.Brandon,F.Armstrong,N.W.Millard,S.D.McPhail,P.Stevenson,M.

Pebody,J.Perrett,andM.Squires.2002.Antarctickrillunderseaice:elevatedabundanceinanarrowbandjust

southoficeedge.Science295:1890-1892.

231


DATE:July2016


Brierley,A.S.,P.G.Fernandes,M.A.Brandon,F.Armstrong,N.W.Millard,S.D.McPhail,P.Stevenson,M.

Pebody,J.Perrett,andM.Squires.2003.AninvestigationofavoidancebyAntarctickrillofRRSJamesClarkRoss

usingtheAutosub-2autonomousunderwatervehicle.FisheriesResearch60:569-576.

Brito,M.,D. Smeed, andG.Griffiths. 2014.Underwater glider reliability and implications for surveydesign.

JournalofAtmosphericandOceanictechnology31:2858-2870.

Buckland,S.T.,D.R.Anderson,K.P.Burnham,J.L.Laake,D.L.Borchers,andL.Thomas.2001.Introductionto

DistanceSampling.Estimatingabundanceofbiologicalpopulations.OxfordUniversityPress,Oxford.

Buckland,S.T.,E.A.Rexstad,T.A.Marques,andC.S.Oedekoven.DistanceSampling:MethodsandApplications.

2015.Springer.

Buckle, J. R., A. Knox, J. Siviter, andA.Montecucco. 2013.AutonomousUnderwaterVehicle Thermoelectric

PowerGeneration.Journalofelectronicmaterials42:2214-2220.

Butler,M.J.A.1988.Theapplicationofremotesensingtechnologytomarinefisheries:anintroductorymanual.

FAOFish.Tech.Pap.,(295):165p.

Caccia,M.Autonomoussurfacecraft:prototypesandbasicresearchissues.null,1-6.2006.IEEE.

Caccia,M.,M.Bibuli,R.Bono,G.Bruzzone,andE.Spirandelli.AluminumhullUSVforcoastalwaterandseafloor

monitoring.OCEANS2009-EUROPE,1-5.2009.IEEE.

Caccia,M.,R.Bono,G.Bruzzone,E.Spirandelli,G.Veruggio,A.M.Stortini,andG.Capodaglio.2005.Sampling

seasurfaceswithSESAMO:anautonomouscraftforthestudyofsea-airinteractions.Robotics&Automation

Magazine,IEEE12:95-105.

Caffaz, A., A. Caiti, G. Casalino, L. Gualdesi, and A. Turetta. 2009. Folaga: a low cost AUV/glider for coastal

environmentalsampling.UnderwaterTechnology28:151-157.

Cameron,M., E.Moreland,andP.Boveng.TestingaShip-BasedUnmannedAerialAircraft System (UAS) for

SurveyingtheBeringSeaPackIce.2009.AFSCQuArterlyreport.

Campbell,S.,W.Naeem,andG.W. Irwin.2012.Areviewon improvingtheautonomyofunmannedsurface

vehiclesthroughintelligentcollisionavoidancemanoeuvres.AnnualReviewsinControl36:267-283.

Campoy,P.,J.F.Correa,I.Mondragon,C.Martinez,M.Olivares,L.Mejias,andJ.Artieda.2009.Computervision

onboardUAVsforciviliantasks.Pages105-135inUnmannedAircraftSystems.Springer.

Cañadas,A.andP.S.Hammond.2006.Model-basedabundanceestimatesforbottlenosedolphinsoffsouthern

Spain:implicationsforconservationandmanagement.JournalofCetaceanResearchandManagement8:13.

Cheney,B.,R.Corkrey,J.W.Durban,K.Grellier,P.S.Hammond,V.Islas-Villanueva,V.M.Janik,S.M.Lusseau,

K.M.Parsons,andN. J.Quick.2014. Long-termtrends in theuseofaprotectedareabysmall cetaceans in

relationtochangesinpopulationstatus.GlobalEcologyandConservation.

232


DATE:July2016


Churnside,J.Ostrovsky,L.andVeenstra,T.2009.ThermalFootprintsofWhales.Oceanography22(1):206-209.

Clark,C.W., Ellison,W.T.,Hatch, L.T. ,Wiley,D.N.,.Merrick,R.L,VanParijs, S.M.2008.AnOceanObserving

System for Large-ScaleMonitoring andMapping ofNoise Throughout the StellwagenBankNationalMarine

Sanctuary.OfficeofNavalResearchScience&Technology.ONRReport(AwardNumber):N00014-07-1-1029.

Claus,B.andR.Bachmayer.2015.Terrain-aidedNavigationforanUnderwaterGlider.JournalofFieldRobotics.

Claus, B. Hybrid glider propulsion module implementation and characterization. Proceedings International

SymposiumonUnmannedUntetheredSubmersibleTechnology,Durham,NH.2009.

Colbo,K.,T.Ross,C.Brown,andT.Weber.2014.Areviewofoceanographicapplicationsofwatercolumndata

frommultibeamechosounders.Estuarine,CoastalandShelfScience145:41-56.

Correa,T.B.,G.P.Eberli,M.Grasmueck,J.K.Reed,andA.M.Correa.2012.Genesisandmorphologyofcold-

watercoralridgesinaunidirectionalcurrentregime.MarineGeology326:14-27.

Cox,M.J.,D.L.Borchers,D.A.Demer,G.R.Cutter,andA.S.Brierley.2011.EstimatingthedensityofAntarctic

krill(Euphausiasuperba)frommulti-beamecho-sounderobservationsusingdistancesamplingmethods.Journal

oftheRoyalStatisticalSociety:SeriesC(AppliedStatistics)60:301-316.

Curé,C.,L.D.Sivle,F.Visser,P.J.Wensveen,S.Isojunno,C.M.Harris,P.H.Kvadsheim,F.P.Lam,andP.J.Miller.

2015.Predatorsoundplaybacksrevealstrongavoidanceresponsesinafightstrategistbaleenwhale.MarEcol

ProgSer526:267-282.

Davis,R.E.,C.C.Eriksen,andC.P.Jones.Autonomousbuoyancy-drivenunderwatergliders.TheTechnologyand

ApplicationsofAutonomousUnderwaterVehicles,37-58.2002.TaylorandFrancis,London.

Davis,R.E.,M.D.Ohman,D.L.Rudnick,andJ.T.Sherman.2008.Glidersurveillanceofphysicsandbiologyin

thesouthernCaliforniaCurrentSystem.LimnologyandOceanography53:2151-2168.

Davy,C.M.andM.B.Fenton.2013.Technicalnote:side-scansonarenablesrapiddetectionofaquaticreptiles

inturbidloticsystems.EuropeanJournalofWildlifeResearch59:123-127.

Dickey,T.D.,E.C.Itsweire,M.A.Moline,andM.J.Perry.2008.IntroductiontotheLimnologyandOceanography

special issue on autonomous and Lagrangian platforms and sensors (ALPS). Limnology and Oceanography

53:2057-2061.

Doray,M.2012.PelagicfishstockassessmentbyacousticmethodsatIfremer.DépartmentÉcologieetModèles

pourl'Halieutique.

D'Spain,G.L.FlyingWingAutonomousUnderwaterGliderforBasicResearchinOceanAcoustics,Signal/Array

Processing,UnderwaterAutonomousVehicle Technology,Oceanography,Geophysics, andMarineBiological

Studies.2009.DTICDocument.

233


DATE:July2016


D'Spain,G.L.,E.Terrill,C.D.Chadwell,J.A.Smith,andR.Zimmerman.2004.Passiveacousticlocalizationwith

anAUV-mountedhydrophonearray.TheJournaloftheAcousticalSocietyofAmerica115:2615.

D'Spain, G. L., J. A. Hildebrand, W. Husband, and M. Stevenson. Follow-On Tests of the ZRay Flying Wing

UnderwaterGliderandWavegliderAutonomousSurfaceVehicles,andtheirPassiveAcousticMarineMammal

MonitoringSystems.2011.

Dudgeon,C.L.,K.H.Pollock,J.M.Braccini,J.M.Semmens,andA.Barnett.2015.Integratingacoustictelemetry

into mark-recapture models to improve the precision of apparent survival and abundance estimates.

Oecologia1-12.

Durban,J.W.,H.Fearnbach,L.G.Barrett-Lennard,W.L.Perryman,andD.J.LeRoi.2015.Photogrammetryof

killerwhalesusingasmallhexacopterlaunchedatsea1.JournalofUnmannedVehicleSystems3:131-135.

Dutton,D.L.,P.H.Dutton,M.Chaloupka,andR.H.Boulon.2005.IncreaseofaCaribbeanleatherbackturtle

Dermochelyscoriaceanestingpopulationlinkedtolong-termnestprotection.BiologicalConservation126:186-

194.

Eiler,J.H.,T.M.Grothues,J.A.Dobarro,andM.M.Masuda.2013.Comparingautonomousunderwatervehicle

(AUV) and vessel-based tracking performance for locating acoustically tagged fish.Marine Fisheries Review

75:27-42.

Eriksen,C.C.,T.J.Osse,R.D.Light,T.Wen,T.W.Lehman,P.L.Sabin,J.W.Ballard,andA.M.Chiodi.2001.

Seaglider:Along-rangeautonomousunderwatervehicleforoceanographicresearch.OceanicEngineering,IEEE

Journalof26:424-436.

FAA.UnmannedAircraftSystems(UAS)OperationalApproval.2013.

Ferguson, S. H., L. Dueck, L. L. Loseto, and S. P. Luque. 2010. BowheadwhaleBalaenamysticetus seasonal

selectionofseaice.MarEcolProgSer411:285-297.

Fernandes,P.G.2009.Classificationtreesforspecies identificationoffish-schoolechotraces. ICESJournalof

MarineScience:JournalduConseilfsp060.

Fernandes,P.G.andA.S.Brierley.1999.Usinganautonomousunderwatervehicleasaplatformformesoscale

acousticsamplinginmarineenvironments.ICESCM1000:01-16.

Fernandes,P.G.,A.S.Brierley,E.J.Simmonds,N.W.Millard,S.D.McPhail,F.Armstrong,P.Stevenson,andM.

Squires.2000.Oceanography:Fishdonotavoidsurveyvessels.Nature404:35-36.

Fernandes,P.G.,P.Stevenson,A.S.Brierley,F.Armstrong,andE.J.Simmonds.2003.Autonomousunderwater

vehicles:futureplatformsforfisheriesacoustics.ICESJournalofMarineScience:JournalduConseil60:684-691.

Flamm,R.O.,E.C.Owen,C.F.Owen,R.S.Wells,andD.Nowacek.2000.AerialVideogrammetryFromATethered

AirshipToAssessManateeLife-stageStructure.MarMamSci16:617-630.

234


DATE:July2016


Foote,K.G.2009.Acousticmethods:briefreviewandprospectsforadvancingfisheriesresearch.Pages313-343

inThefutureoffisheriesscienceinNorthAmerica.Springer.

Fotinopoulos, V. Balloon photogrammetry for archaeological surveys. International Archives of the

Photogrammetry,RemoteSensingandSpatialInformationSciences,XXISPRSCongress,Istanbul,Turkey35,B5.

2004.

Fraser,W.R.,J.C.Carlson,P.A.Duley,E.J.Holm,andD.L.Patterson.1999.Usingkite-basedaerialphotography

forconductingAdeliepenguincensusesinAntarctica.Waterbirds435-440.

Fratantoni,D.M.,M.F.Baumgartner,andM.P.Johnson.DevelopmentandValidationofaMobile,Autonomous,

BroadbandPassiveAcousticMonitoringSystemforMarineMammals.2011.DTICDocument.

Fucile,P.D.,R.C.Singer,M.Baumgartner,andK.Ball.ASelfContainedRecorderForAcousticObservationsfrom

AUV's.OCEANS2006,1-4.2006.IEEE.

Funaki,M.andN.Hirasawa.2008.Outlineofasmallunmannedaerialvehicle(Ant-Plane)designedforAntarctic

research.PolarScience2:129-142.

Garner,G.,S.Amstrup,J.Laake,B.Manley,L.McDonald,andD.Robertson.1999.Marinemammalsurveyand

assessmentmethods.CRCPress.

Gassmann,M.,S.M.Wiggins,andJ.A.Hildebrand.2015.Three-dimensionaltrackingofCuvier'sbeakedwhales'

echolocation sounds using nested hydrophone arrays. The Journal of the Acoustical Society of America

138:2483-2494.

Gibbens,D.2014.IntegratingUASintotheoilandgasindustry.JournalofUnmannedVehicleSystems2:vi-vii.

Glennie,R.,S.T.Buckland,andL.Thomas.2015.Theeffectofanimalmovementonlinetransectestimatesof

abundance.PLoSONE10:e0121333.

Goebel,M.E.,W.L.Perryman,J.T.Hinke,D.J.Krause,N.A.Hann,S.Gardner,andD.J.LeRoi.2015.Asmall

unmannedaerialsystemforestimatingabundanceandsizeofAntarcticpredators.PolarBiology38:619-630.

Gonzalez-Socoloske,D.andL.D.Olivera-Gomez.2012.Gentlegiantsindarkwaters:usingside-scansonarfor

manateeresearch.TheOpenRemoteSensingJournal5:1-14.

Gonzalez-Socoloske, D., L. D. Olivera-Gomez, and R. E. Ford. 2009. Detection of free-ranging West Indian

manateesTrichechusmanatususingside-scansonar.EndangeredSpeciesResearch8:249-257.

Gordon,J.,D.Gillespie,J.Potter,A.Frantzis,M.P.Simmonds,R.Swift,andD.Thompson.2003.Areviewofthe

effectsofseismicsurveysonmarinemammals.MarineTechnologySocietyJournal37:16-34.

Goudey,C.A.,T.Consi,J.Manley,andM.Graham.1998.Aroboticboatforautonomousfishtracking.Marine

TechnologySociety.MarineTechnologySocietyJournal32:47.

235


DATE:July2016


Grasmueck,M.,G.P.Eberli,D.A.Viggiano,T.Correa,G.Rathwell,andJ.Luo.2006.Autonomousunderwater

vehicle(AUV)mappingrevealscoralmounddistribution,morphology,andoceanographyindeepwaterofthe

StraitsofFlorida.GeophysicalResearchLetters33.

Graver,J.G.2005.Underwatergliders:Dynamics,controlanddesign.PrincetonUniversityNewJersey,USA.

Greene,C.H.,E.L.Meyer-Gutbrod,L.P.McGarry,L.C.HufnagleJr,D.Chu,S.McClatchie,A.Packer,J.B.Jung,

T.Acker,andH.Dorn.2014.Awavegliderapproachtofisheriesacoustics:Transforminghowwemonitorthe

nation'scommercialfisheriesinthe21stcentury.Oceanography27:168-174.

Greene,C.H.,E.L.Meyer-Gutbrod,L.P.McGarry,L.C.HufnagleJr,D.Chu,S.McClatchie,A.Packer,J.B.Jung,

T.Acker,andH.Dorn.2014.Awavegliderapproachtofisheriesacoustics:Transforminghowwemonitorthe

nation'scommercialfisheriesinthe21stcentury.Oceanography27:168-174.

Grenzdörffer,G.+.,A.N.G.E.Engel,andK.Jütte.2008.Low-costUAV'sinderLand-undForstwirtschaft-Eine

vergleichendeUntersuchungvonzweiMini-UAV's.Seyfert,E.,DGPF-Jahrestagung27-36.

Griffiths,E.T.andJ.Barlow.2015.EquipmentPerformanceReportfortheDriftingAcousticSparBuoyRecorder

(DASBR).

Griffiths,E.T.andJ.Barlow.2015.EquipmentPerformanceReportfortheDriftingAcousticSparBuoyRecorder

(DASBR).

Griffiths,G.2002.Technologyandapplicationsofautonomousunderwatervehicles.2ed.CRCPress.

Griffiths,G.,C.Jones,J.Ferguson,andN.Bose.2007.Underseagliders.JournalofOceanTechnology2:64-75.

Grimm, B. A., B. A. Lahneman, P. B. Cathcart, R. C. Elgin, G. L.Meshnik, and J. P. Parmigiani. Autonomous

unmanned aerial vehicle system for controlling pest bird population in vineyards. ASME 2012 International

MechanicalEngineeringCongressandExposition,499-505.2012.AmericanSocietyofMechanicalEngineers.

GROOM. Gliders for Research, Ocean Observation and Management FP7-Infra-2011-2.1.1 "Design Studies"

DeliverableD2.3LegalaspectsofglideroperationsinEuropeanWaters.2013.

Guihen,D.,S.Fielding,E.J.Murphy,K.J.Heywood,andG.Griffiths.2014.Anassessmentoftheuseofocean

gliders to undertake acoustic measurements of zooplankton: the distribution and density of Antarctic krill

(Euphausiasuperba)intheWeddellSea.LimnologyandOceanography-Methods12:373-389.

Harris,D.,L.Thomas,D.Gillespie,andT.Yack.Effectsofcetaceandepthonacousticdistancesamplingsurveys.

Bookofabstracts,94.2014.

Hastie, G. D., C. Donovan, T. Goetz, andV.M. Janik. 2014. Behavioral responses by grey seals (Halichoerus

grypus)tohighfrequencysonar.MarinePollutionBulletin79:205-210.

Hastie,G.TrackingMarineMammalsAroundMarineRenewableEnergyDevicesUsingActiveSonar.SMRULtd

reportURN:12D/328totheDepartmentofEnergyandClimateChange.2012.

236


DATE:July2016


Hastie,G.,C.Sparling,andA.Murray.Sonarasatooltomonitorinteractionsbetweenmarinemammalsand

tidalturbines;pitfallsandpossibilities.EnvironmentalInteractionsofMarineRenewableEnergyTechnologies.

2012.

Haulsee,D.E.,M.W.Breece,D.C.Miller,B.M.Wetherbee,D.A.Fox,andM.J.Oliver.2015.Habitatselection

ofacoastalsharkspeciesestimatedfromanautonomousunderwatervehicle.

Heywood,K.J.,S.Schmidtko,C.+.Heuzé,J.Kaiser,T.D.Jickells,B.Y.Queste,D.P.Stevens,M.Wadley,A.F.

Thompson,andS.Fielding.2014.OceanprocessesattheAntarcticcontinentalslope.PhilosophicalTransactions

oftheRoyalSocietyofLondonA:Mathematical,PhysicalandEngineeringSciences372:20130047.

Hilborn,R.andC.J.Walters.1992.Quantitativefisheriesstockassessment:choice,dynamicsanduncertainty.

ReviewsinFishBiologyandFisheries2:177-178.

Hildebrand,J.A.,S.Baumann-Pickering,K.E.Frasier,J.S.Trickey,K.P.Merkens,S.M.Wiggins,M.A.McDonald,

L.P.Garrison,D.Harris,andT.A.Marques.2015.Passiveacousticmonitoringofbeakedwhaledensitiesinthe

GulfofMexico.ScientificReports5.

Hodgson,A.2007.BLIMP-CAM:AerialVideoObservationsofMarineAnimals.MarineTechnologySocietyJournal

41:39-43.

Hodgson,A. J.,M.Noad,H.Marsh, J. Lanyon,andE.Kniest.UsingUnmannedAerialVehicles for surveysof

marinemammals in Australia: test of concept. 2010. Final report to the australianmarinemammal centre,

UniversityofQueenslandandJamesCookUniversityandUniversityofNewcastle,Australian.

Hodgson,A.,N.Kelly,andD.Peel.2013.UnmannedAerialVehicles(UAVs)forSurveyingMarineFauna:ADugong

CaseStudy.PLoSONE8.

Horne,J.K.2000.Acousticapproachestoremotespeciesidentification:areview.FisheriesOceanography9:356-

371.

Huse,G.,B.R.MacKenzie,V.Trenkel,M.Doray,L.Nøttestad,andG.Oskarsson.2015.Spatiallyexplicitestimates

ofstocksizes,structureandbiomassofherringandbluewhiting,andcatchdataofbluefintuna.EarthSystem

ScienceData7:35-46.

Imlach,J.andR.Mahr.Modificationofamilitarygradegliderforcoastalscientificapplications.Oceans,2012,

1-6.2012.IEEE.

Inoue,J.,J.A.Curry,andJ.A.Maslanik.2008.Applicationofaerosondestomelt-pondobservationsoverArctic

seaice.JournalofAtmosphericandOceanictechnology25:327-334.

Ireland,D.,K.Leonard,G.Schaefer,C.Sparks,R. Jannarone,B.Koski,D.Funk,A.Macrander,andK.Broker.

AutomatedDetectionofLargeCetaceansinAerialDigitalImagery.Papertobepresentedattheannualmeeting

237


DATE:July2016


of21stBiennial InternationalConferenceon theBiologyofMarineMammals:Bridging thePastToward the

Future,SanFrancisco,CAUSA.2015.

Johannesson,K.A.andR.B.Mitson.1983.Fisheriesacoustics:apracticalmanualforaquaticbiomassestimation.

FoodandAgricultureOrganizationoftheUnitedNations.

Johnson,M.,D.Fratantoni,M.Baumgartner,andT.Hurst.BetaTestingofPersistentPassiveAcousticMonitors.

2012.DTICDocument.

Jones,C.,D.Webb,S.Glenn,O.Schofield,J.Kerfoot,J.Kohut,D.Aragon,C.Haldeman,T.Haskin,andA.Kahl.

Slocumgliderexpandingthecapabilities.Proc.17thInt.Symp.UnmannedUntetheredSubmers.Technol.Uust11.

2011a.

Jones,D.W.,R.J.Carr,andB.Schwartz-Gilbert.GLMPC:Anintegratedsystemforglidercommandandcontrol.

OCEANS2011,1-7.2011b.IEEE.

Jones,I.,L.G.Pearlstine,andH.F.Percival.2006.Anassessmentofsmallunmannedaerialvehiclesforwildlife

research.WildlifeSocietyBulletin34:750-758.

Kaatz, I. M. 2002. Multiple sound-producing mechanisms in teleost fishes and hypotheses regarding their

behaviouralsignificance.Bioacoustics12:230-233.

Karniski,C.,E.M.Patterson,E.Krzyszczyk,V.Foroughirad,M.A.Stanton,andJ.Mann.2014.Acomparisonof

surveyandfocalfollowmethodsforestimatingindividualactivitybudgetsofcetaceans.MarMamSci.

Kasumyan,A.O.2008.Soundsandsoundproductioninfishes.JournalofIchthyology48:981-1030.

Kaufman,D.E.,M.A.Friedrichs,W.O.Smith,B.Y.Queste,andK.J.Heywood.2014.Biogeochemicalvariability

inthesouthernRossSeaasobservedbyagliderdeployment.DeepSeaResearchPartI:OceanographicResearch

Papers92:93-106.

Kennish,M.J.,S.M.Haag,G.P.Sakowicz,andR.A.Tidd.2004.Side-ScanSonar ImagingofSubtidalBenthic

HabitatsintheMullicaRiver-GreatBayEstuarineSystem.JournalofCoastalResearch227-240.

Ketten,D.R.1995.Estimatesofblastinjuryandacoustictraumazonesformarinemammalsfromunderwater

explosions.Sensorysystemsofaquaticmammals391-407.

Klinck,H.,D.K.Mellinger,K.Klinck,N.M.Bogue,J.C.Luby,W.A.Jump,G.B.Shilling,T.Litchendorf,A.S.Wood,

G.S.Schorr,andR.W.Baird.2012.Near-Real-TimeAcousticMonitoringofBeakedWhalesandOtherCetaceans

UsingaSeaglider(TM).PLoSONE7.

Klinck, H., H. Matsumoto, S. Fregosi, and D. K. Mellinger. 2014. High-frequency observations from mobile

autonomousplatforms.TheJournaloftheAcousticalSocietyofAmerica136:2119.

Klinck,H.,K.Stelzer,K.Jafarmadar,andD.K.Mellinger.Aasendurance:Anautonomousacousticsailboatfor

marinemammalresearch.2009.

238


DATE:July2016


Knudsen,F.,O.Gammelsæter,P.H.Kvadsheim,andL.Nøttestad.2008.Evaluationoffisheriessonarforwhale

detectioninrelationtoseismicsurveyoperations.Bioacoustics17:326-328.

Kobayashi,D.R.,J.J.Polovina,D.M.Parker,N.Kamezaki,I.J.Cheng,I.Uchida,P.H.Dutton,andG.H.Balazs.

2008. Pelagic habitat characterizationof loggerhead sea turtles,Caretta caretta, in theNorth PacificOcean

(1997-2006): insights from satellite tag tracking and remotely sensed data. Journal of ExperimentalMarine

BiologyandEcology356:96-114.

Koh,L.andS.Wich.2012.Dawnofdroneecology:low-costautonomousaerialvehiclesforconservation.Tropical

ConservationScience5:121-132.

Korneliussen,R.J.andE.Ona.2003.Syntheticechogramsgeneratedfromtherelativefrequencyresponse.ICES

JournalofMarineScience:JournalduConseil60:636-640.

Korneliussen,R.J.,Y.Heggelund,I.K.Eliassen,O.K.Øye,T.Knutsen,andJ.Dalen.2009.Combiningmultibeam-

sonarandmultifrequency-echosounderdata:examplesoftheanalysisandimagingoflargeeuphausiidschools.

ICESJournalofMarineScience:JournalduConseilfsp092.

Koski,W.R.,P.Abgrall, andS.B.Yazvenko.2009a.AReviewand inventoryofunmannedaerial systems for

detection and monitoring of key biological resources and physical parameters affecting marine life during

offshoreexplorationandproductionactivities.IWCpaperSC/61E9.

Koski,W.R.,T.A.Thomas,D.W.Funk,andA.M.Macrander.2013.Marinemammalsightingsbyanalystsof

digitalimageryversusaerialsurveyors:apreliminarycomparison.JournalofUnmannedVehicleSystems1:25-

40.

Koski,W.R.,T.Allen,D.Ireland,G.Buck,P.R.Smith,A.M.Macrander,M.A.Halick,C.Rushing,D.J.Sliwa,and

T.L.McDONALD.2009b.Evaluationofanunmannedairbornesystemformonitoringmarinemammals.Aquatic

Mammals35:347.

Kraska,J.2010.Thelawofunmannednavalsystemsinwarandpeace.JournalofOceanTechnology.

Kusel,E.T.,D.K.Mellinger,L.Thomas,T.A.Marques,D.Moretti,andJ.Ward.2011.Cetaceanpopulationdensity

estimation from single fixed sensors using passive acoustics. Journal of the Acoustical Society of America

129:3610-3622.

Leonard,N.E.,D.A.Paley,R.E.Davis,D.M.Fratantoni,F.Lekien,andF.Zhang.2010.Coordinatedcontrolofan

underwater glider fleet in an adaptive ocean sampling field experiment in Monterey Bay. Journal of Field

Robotics27:718-740.

Lewis,T.,D.Gillespie,C.Lacey,J.Matthews,M.Danbolt,R.Leaper,R.Mclanaghan,andA.Moscrop.2007.Sperm

whaleabundanceestimatesfromacousticsurveysoftheIonianSeaandStraitsofSicilyin2003.Journalofthe

MarineBiologicalAssociationoftheUnitedKingdom87:353-357.

239


DATE:July2016


Lieber,L.,B.Williamson,C.S.Jones,L.R.Noble,A.Brierley,P.Miller,andB.E.Scott.Introducingnovelusesof

multibeamsonartostudybaskingsharksinthelightofmarinerenewableenergyextraction.Proceedingsofthe

2ndInternationalConferenceonEnvironmentalInteractionsofMarineRenewableEnergyTechnologies.2014.

Loe,Ø.A.2008.CollisionAvoidanceforUnmannedSurfaceVehicles.MScNorwegianUniversityofScienceand

Technology.

Love,R.H.1973.TargetstrengthsofhumpbackwhalesMegapteranovaeangliae.TheJournaloftheAcoustical

SocietyofAmerica54:1312-1315.

Lucieer, V.,N.A.Hill,N. S. Barrett, and S.Nichol. 2013.Domarine substrates 'look' and 'sound' the same?

Supervisedclassificationofmultibeamacousticdatausingautonomousunderwatervehicleimages.Estuarine,

CoastalandShelfScience117:94-106.

Lucifredi,I.andP.J.Stein.2007.Graywhaletargetstrengthmeasurementsandtheanalysisofthebackscattered

response.TheJournaloftheAcousticalSocietyofAmerica121:1383-1391.

Lucke,K.,P.A.Lepper,M.A.Blanchet,andU.Siebert.2008.Testingtheacoustictoleranceofharbourporpoise

hearingforimpulsivesounds.Bioacoustics17:329-331.

Luftfartstilsynet.IOGPUnmannedAerialSystemsGuidelines.HøringsinnspilltilhøringavRPAS-forskrift–Statoil

ASA. Available at: http://www.luftfartstilsynet.no/incoming/Statoil_ASA_-

_vedlegg_til_h%C3%B8ringsinnspill_IOGP_UAS_Guidelines_V1_Feb_2015.pdf/BINARY/Statoil%20ASA%20-

%20vedlegg%20til%20h%C3%B8ringsinnspill%20(IOGP%20UAS%20Guidelines%20V1%20Feb%202015).pdf.

2015.

Lyons,C.,W.R.Koski,andD.S. Ireland.2006.Unmannedaerial surveys.(Chapter8,15pp). JointMonitoring

ProgramintheChukchiandBeaufortseas,openwaterseasons2008.

MacKellar,M.C.,H.A.McGowan,andS.R.Phinn.2013.Anobservationalheatbudgetanalysisofacoralreef,

HeronReef,GreatBarrierReef,Australia.JournalofGeophysicalResearch:Atmospheres118:2547-2559.

MacLennan,D.N. andE. J. Simmonds. 1992. Fish andFisheries Series 5. Fisheries acousticChapmanyHall,

London.

Mahfurdz,A.andA.Hamzah.AcousticStrengthComparisonbetweenGreenTurtleandFishUsingEchoSounder.

AnnualConferenceonEngineeringandInformation,28-30March2014,Tokyo,Jepun.2014.

Mahfurdz,A.,A.H.Sunardi,S.A.S.A.Kadir,andN.Sulong.2013.AcousticStrengthofGreenTurtleandFish

BasedonFFTAnalysis.InternationalJournalofAdvancedComputerScienceandApplications9:74-78.

Mahfurdz,A.,S.Sunardi,H.Ahmad,S.Abdullah,andN.Nazuki.2015.DistinguishSeaTurtleandFishUsingSound

Technique inDesigningAcousticDeterrentDevice. TELKOMNIKA (TelecommunicationComputing Electronics

andControl)13:1305-1311.

240


DATE:July2016


Maldague, X. Multipolar Infrared Vision. Infrared Vision - Electromagnetic Spectrum (Chapter 1). Laval

University,DepartmentofElectricalandComputerEngineering.2007.

Manley,J.E.Unmannedsurfacevehicles,15yearsofdevelopment.OCEANS2008,1-4.2008.IEEE.

Manley,J.E.andT.W.Vaneck.1998.Highfidelityhydrographicsurveysusingandautonomoussurfacecraft.

Manley,J.E.Unmannedsurfacevehicles,15yearsofdevelopment.OCEANS2008,1-4.2008.IEEE.

Manley, J.E.,A.Marsh,W.Cornforth,andC.Wiseman.Evolutionof theautonomoussurfacecraftAutoCat.

Oceans2000MTS/IEEEConferenceandExhibition1,403-408.2000.IEEE.

Marques,T.A.,L.Munger,L.Thomas,S.Wiggins,andJ.A.Hildebrand.2011.EstimatingNorthPacificrightwhale

Eubalaenajaponicadensityusingpassiveacousticcuecounting.EndangeredSpeciesResearch13:163-172.

Marques,T.A.,L.Thomas,J.Ward,D.Nancy,andP.L.Tyack.2009.Estimatingcetaceanpopulationdensityusing

fixedpassiveacousticsensors:AnexamplewithBlainville'sbeakedwhales.JournaloftheAcousticalSocietyof

America125:1982-1994.

Marques,T.A.,L.Thomas,S.W.Martin,D.K.Mellinger,J.A.Ward,D.J.Moretti,D.Harris,andP.L.Tyack.2013.

Estimatinganimalpopulationdensityusingpassiveacoustics.BiologicalReviews88:287-309.

Marques,T.A.,S.T.Buckland,R.Bispo,andB.Howland.2013a.Accountingforanimaldensitygradientsusing

independentinformationindistancesamplingsurveys.StatisticalMethodsandApplications22:67-80.

Martin,J.,H.H.Edwards,M.A.Burgess,H.Percival,D.E.Fagan,B.E.Gardner,J.G.Ortega-Ortiz,P.G.Ifju,B.S.

Evers, and T. J. Rambo. 2012. Estimating Distribution of Hidden Objects with Drones: From Tennis Balls to

Manatees.PLoSONE7.

Matsumoto,H.,C.Jones,H.Klinck,D.K.Mellinger,R.P.Dziak,andC.Meinig.2013.Trackingbeakedwhaleswith

apassiveacousticprofilerfloat.JournaloftheAcousticalSocietyofAmerica133:731-740.

Mayer,L.A.2006.Frontiersinseafloormappingandvisualization.MarineGeophysicalResearches27:7-17.

Mejias, L.,G.Duclos,A.Hodgson, andF.Maire.Automatedmarinemammaldetection fromaerial imagery.

Oceans-SanDiego,2013,1-5.2013.IEEE.

Mellody,M. 2015. RobustMethods for the Analysis of Images and Videos for Fisheries Stock Assessment:

SummaryofaWorkshop.NationalAcademiesPress.

Miles,T.,G.Seroka,J.Kohut,O.Schofield,andS.Glenn.2015.Gliderobservationsandmodelingofsediment

transportinHurricaneSandy.JournalofGeophysicalResearch:Oceans120:1771-1791.

Miles,T.,S.M.Glenn,andO.Schofield.2013.Temporalandspatialvariabilityinfallstorminducedsediment

resuspensionontheMid-AtlanticBight.ContinentalShelfResearch63:S36-S49.

241


DATE:July2016


Miller,D.L.,M.L.Burt,E.A.Rexstad,andL.Thomas.2013.Spatialmodelsfordistancesamplingdata:recent

developmentsandfuturedirections.MethodsEcolEvol4:1001-1010.

Miller,J.H.andD.C.Potter.Activehighfrequencyphased-arraysonarforwhaleshipstrikeavoidance:Target

strengthmeasurements.OCEANS,2001.MTS/IEEEConferenceandExhibition4,2104-2107.2001.IEEE.

Miller,P.J.,R.N.Antunes,P.J.Wensveen,F.I.Samarra,A.C.Alves,P.L.Tyack,P.H.Kvadsheim,L.Kleivane,F.

P.Lam,M.A.Ainslie,andL.Thomas.2014.Dose-responserelationshipsfortheonsetofavoidanceofsonarby

free-rangingkillerwhales.JournaloftheAcousticalSocietyofAmerica135:975-993.

Moreland,E.E.,M.F.Cameron,R.P.Angliss,andP.L.Boveng.2015.Evaluationofaship-basedunoccupied

aircraftsystem(UAS)forsurveysofspottedandribbonsealsintheBeringSeapackice1.JournalofUnmanned

VehicleSystems3:114-122.

Mulaca,B.,R.Storvoldb,andE.Weatherhead.RemotesensingintheArcticwithunmannedaircraft:Helping

scientists to achieve their goals. Proceedings of 34th International Symposium on Remote Sensing of

Environment,Sydney,Australia,10-15.2011.

Mulero-Pázmány,M.,R. Stolper, L.D.VanEssen, J. J.Negro, andT. Sassen. 2014.Remotelypilotedaircraft

systemsasarhinocerosanti-poachingtoolinAfrica.PLoSONE9:e83873.

Neininger,B.andJ.M.Hacker.Mannedorunmanned?doesthisreallymatter. InternationalArchivesof the

Photogrammetry, Remote Sensing and Spatial Information Sciences, Zurich, Switzerland.Conference on

UnmannedAerialVehicleinGeomatics38,1.2011.

NMFS. Supplemental Environmental Assessment On The Effects of the Issuance of Eleven NationalMarine

FisheriesService(NMFS)PermittedScientificResearchActivitiesonMarineMammalsandSeaTurtlesintheU.S.

TerritorialWatersandHighSeasoftheNorthPacificOcean(IncludingtheGulfofAlaskaandBeringSea),Arctic

Ocean (Including theChukchi SeaandBeaufort Sea), SouthernOcean (IncludingWatersoffAntarctica), and

ForeignTerritorialWatersofMexico(GulfofCaliforniaOnly),Canada,Russia,Japan,andthePhilippines.2008.

NOAA.2015.WhatisLIDAR?NationalOceanicandAtmosphericAdministration.NationalOceanService.

Nøttestad,L.,A.Fernö,S.Mackinson,T.Pitcher,andO.A.Misund.2002.Howwhalesinfluenceherringschool

dynamicsinacold-frontareaoftheNorwegianSea.ICESJournalofMarineScience59:393-400.

NSF. 2009. Ocean Observatories Initiative - Construction under MREFC. NSF Program: Ocean Observ Init-

Construction.ProgramReferenceCode(s):0000,OTHR,037F.ProgramElementCode:039F.

Ohman,M.D.,D.L.Rudnick,A.M.Chekalyuk,R.E.Davis,R.A.Feely,M.Kahru,H.J.Kim,M.R.Landry,T.R.

Martz, and C. L. Sabine. 2013. Autonomous ocean measurements in the California Current Ecosystem.

Oceanography26:18-25.

242


DATE:July2016


Olivares-Mendez,M.A.,T.F.Bissyandé,K.Somasundar, J.Klein,H.Voos,andY. LeTraon.2013.TheNOAH

project:Givingachancetothreatenedspecies inAfricawithUAVs.Pages198-208 ine-Infrastructureande-

ServicesforDevelopingCountries.Springer.

Osse,T.J.andC.C.Eriksen.TheDeepglider:Afulloceandepthgliderforoceanographicresearch.OCEANS2007

,1-12.2007.IEEE.

Parsons,M.J.,I.M.Parnum,K.Allen,R.D.McCauley,andC.Erbe.2014.DetectionofsharkswiththeGemini

imagingsonar.AcousticsAustralia42:185.

Pascoal,A.,P.Oliveira,C.Silvestre,L.Sebastiao,M.Rufino,V.Barroso,J.Gomes,G.Ayela,P.Coince,andM.

Cardew.Roboticoceanvehicles formarine scienceapplications: theeuropeanasimovproject.Oceans2000

MTS/IEEEConferenceandExhibition1,409-415.2000.IEEE.

Patel,R.2004.Remotelycontrollingwindowsapplications-Autonomousvehiclesstillneedadriver.DRDOBBS

JOURNAL29:14-+.

Pérez-Arjona,I.,V.Espinosa,E.Alonso,P.Ordóñez,S.Llorens,V.Puig,M.Rodilla,J.Castaño,J.A.Esteban,and

J.Eymar.TSmeasurementsandsimulationsofMediterraneanseaturtles.RolfJ.Korneliussen.Proceedingsof

the38thScandinavianSymposiumonPhysicalAcoustics,Geilo30February2-6,2013.2013.

Petitgas,P.,J.Massé,P.Beillois,E.Lebarbier,andA.LeCann.2003.Samplingvarianceofspeciesidentification

infisheriesacousticsurveysbasedonautomatedproceduresassociatingacousticimagesandtrawlhauls.ICES

JournalofMarineScience:JournalduConseil60:437-445.

Pirotta,E.,K.L.Brookes, I.M.Graham,andP.M.Thompson.2014.Variation inharbourporpoiseactivity in

responsetoseismicsurveynoise.Biologyletters10:20131090.

Pirotta,E.,L.New,J.Harwood,andD.Lusseau.2014.Activities,motivationsanddisturbance:Anagent-based

model of bottlenose dolphin behavioral dynamics and interactions with tourism in Doubtful Sound, New

Zealand.EcologicalModelling282:44-58.

Piterbarg,L.,V.Taillandier,andA.Griffa.2014.Investigatingfrontalvariabilityfromrepeatedglidertransectsin

theLigurianCurrent(NorthWestMediterraneanSea).JournalofMarineSystems129:381-395.

Planck, M. K. E. L. 1914. The theory of heat radiation, translated byM.Masius. P. Blackiston's Son & Co,

Philadelphia,reprintedbyKessinger.

Pomeroy,P.,L.O'Connor,andP.Davies.2015.Assessinguseofandreactiontounmannedaerialsystemsingray

andharborsealsduringbreedingandmoltintheUK.JournalofUnmannedVehicleSystems3:102-113.

Powell, J. R. and M. D. Ohman. 2015a. Changes in zooplankton habitat, behavior, and acoustic scattering

characteristics across glider-resolved fronts in the Southern California Current System. Progress in

Oceanography134:77-92.

243


DATE:July2016


Powell,J.R.andM.D.Ohman.2015b.Covariabilityofzooplanktongradientswithglider-detecteddensityfronts

intheSouthernCaliforniaCurrentSystem.DeepSeaResearchPartII:TopicalStudiesinOceanography112:79-

90.

Pyc,C.D.,M.Geoffroy, andF.R.Knudsen.2015.Anevaluationof activeacousticmethods fordetectionof

marinemammalsintheCanadianBeaufortSea.MarMamSci.

Rochet,V.Multi-missionsSeaExplorerglider.ALSEAMARALCENInnovationatSea.2015.

Rogers,R.ThelegalregimegoverningAUVoperationshowfarandhowwide.AutonomousUnderwaterVehicles

(AUV),2012IEEE/OES,1-5.2012.IEEE.

Rountree,R.A.,R.G.Gilmore,C.A.Goudey,A.D.Hawkins,J.J.Luczkovich,andD.A.Mann.2006.Listeningto

fish:applicationsofpassiveacousticstofisheriesscience.Fisheries31:433-446.

Rowat,D.,M.Gore,M.G.Meekan,I.R.Lawler,andC.J.Bradshaw.2009.Aerialsurveyasatooltoestimate

whalesharkabundancetrends.JournalofExperimentalMarineBiologyandEcology368:1-8.

Rudnick,D.L.,R.E.Davis,C.C.Eriksen,D.M.Fratantoni,andM.J.Perry.2004.Underwaterglidersforocean

research.MarineTechnologySocietyJournal38:73-84.

Sagen,H.,S.Sandven,A.Beszczynska-Möller,O.Boebel,T.F.Duda,L.Freitag,J.C.Gascard,A.Gavrilov,C.M.

Lee,andD.K.Mellinger.Acoustictechnologiesforobservingthe interioroftheArcticOcean.OceanObs'09:

SustainedOceanObservationsandInformationforSociety,Venice,Italy.2009.

Sagen,H.,S.Sandven,P.Worcester,M.Dzieciuch,andE.Skarsoulis.2008.TheFramStraitacoustictomography

system.JournaloftheAcousticalSocietyofAmerica123:2991.

Santhaseelan,V.andV.K.Asari.2015.AutomatedWhaleBlowDetectioninInfraredVideo.ComputerVision

andPatternRecognitioninEnvironmentalInformatics58.

Scheritz,M., R. Dietrich, S. Scheller,W. Schneider, and J. Boike. High Resolution Digital ElevationModel of

Polygonal Patterned Ground on Samoylov Island, Siberia, Using Small-Format Photography. United states

permafrostassociation,9thInternationalConferenceonPermafrost,UniversityofAlaska,Fairbanks,USA,1589-

1594.2008.

Schofield,O.,J.Kohut,D.Aragon,L.Creed,J.Graver,C.Haldeman,J.Kerfoot,H.Roarty,C.Jones,andD.Webb.

2007.Slocumgliders:Robustandready.JournalofFieldRobotics24:473-486.

Schofield,O.,J.Kohut,S.Glenn,J.Morell,J.Capella,J.Corredor,J.Orcutt,M.Arrott,I.Krueger,andM.Meisinger.

2010. A regional Slocum glider network in theMid-Atlantic Bight leverages broad community engagement.

MarineTechnologySocietyJournal44:185-195.

244


DATE:July2016


Scott-Hayward,L.,X.S.Oedekoven,M.L.Mackenzie,C.G.Walker,andE.Rexstad.2013.Userguideforthe

MRSeapackage.Statisticalmodellingofbirdandcetaceandistributions inoffshorerenewablesdevelopment

areas.

Seber,G.A.F.1982.Theestimationofanimalabundance.GriffinLondon.

Selivanovsky, D. A. and A. B. Ezersky. 1996. Sound scattering by hydrodynamic wakes of sea animals.

OceanographicLiteratureReview11:1180.

Sherman,J.,R.E.Davis,W.B.Owens,andJ.Valdes.2001.Theautonomousunderwaterglider"Spray".Oceanic

Engineering,IEEEJournalof26:437-446.

Simmonds,E.J.andD.N.MacLennan.Fisheriesacoustics:Theoryandpractise.Oxford:BlackwellScience;2005.

2005.

Smeed,D.,G.McCarthy,andD.White.Underwaterglidersasvirtualmoorings;lessonsfromtheRAPIDprogram.

EGUGeneralAssemblyConferenceAbstracts15,6350.2013.

Smith,C.,S.T.Sykora-Bodie,B.Bloodworth,S.M.Pack,T.R.Spradlin,andN.R.LeBoeuf.2016.Assessmentof

knownimpactsofunmannedaerialsystems(UAS)onmarinemammals:datagapsandrecommendationsfor

researchersintheUnitedStates1.JournalofUnmannedVehicleSystems4:1-14.

Smolowitz,R.J.,S.H.Patel,H.L.Haas,andS.A.Miller.2015.Usingaremotelyoperatedvehicle(ROV)toobserve

loggerheadseaturtle(Carettacaretta)behavioronforaginggroundsoffthemid-AtlanticUnitedStates.Journal

ofExperimentalMarineBiologyandEcology471:84-91.

Sonnenburg, C. R. and C. A.Woolsey. 2013.Modeling, identification, and control of an unmanned surface

vehicle.JournalofFieldRobotics30:371-398.

Sousa-Lima, R. S., T. F. Norris, J. N. Oswald, and D. P. Fernandes. 2013. A Review and Inventory of Fixed

AutonomousRecordersforPassiveAcousticMonitoringofMarineMammals.AquaticMammals39:23-53.

Sparling,C.,D.Gillespie,G.Hastie,J.Gordon,J.Macaulay,C.Malinka,M.Wu,andMcConnell,B.2016.Scottish

GovernmentDemonstrationStrategy:TriallingMethodsforTrackingtheFineScaleUnderwaterMovementsof

MarineMammalsinAreasofMarineRenewableEnergyDevelopment.PublishedbyMarineScotlandScience.

ScottishMarineandFreshwaterScienceVol7No14.

Stadler,M.,M.Salmon,andC.Roberts.2014.EcologicalCorrelatesofGreenTurtle(Cheloniamydas)Abundance

ontheNearshoreWormReefsofSoutheasternFlorida.JournalofCoastalResearch31:244-254.

Stevenson, B. C., D. L. Borchers, R. Altwegg, R. J. Swift, D.M. Gillespie, and G. J.Measey. 2015. A general

frameworkforanimaldensityestimationfromacousticdetectionsacrossafixedmicrophonearray.Methods

EcolEvol6:38-48.

Stommel,H.1989.TheSLOCUMmission.Oceanography2:22-25.

245


DATE:July2016


Suberg, L., R. B. Wynn, J. van der Kooij, L. Fernand, S. Fielding, D. Guihen, D. Gillespie, M. Johnson, K. C.

Gkikopoulou, and I. J. Allan. 2014. Assessing the potential of autonomous submarine gliders for ecosystem

monitoringacrossmultipletrophiclevels(planktontocetaceans)andpollutantsinshallowshelfseas.Methods

inOceanography10:70-89.

Thamm, H. P. 2011. SUSI62 a robust and safe parachute UAV with long flight time and good payload.

InternationalArchivesofthePhotogrammetry,RemoteSensingandSpatialInformationSciences38:1.

Thamm,H.P.andM.Judex.Thelowcostdrone aninterestingtoolforprocessmonitoringinahighspatialand

temporal resolution. The International Archives of the Photogrammetry, Remote Sensing and Spatial

InformationSciences,ISPRSCommissionVIIMid-termSymposium"RemoteSensing:FromPixelstoProcesses",

Enschede,theNetherlands,XXXVIpart7,140-144.ISPRSMid-termSymposium,8-11.2006.

Thamm,H. P., T. Ludwig, and C. Reuter. Design of a ProcessModel forUnmannedAerial Systems (UAS) in

Emergencies.ProceedingsoftheInformationSystemsforCrisisResponseandManagement(ISCRAM),478-487.

2013.

Thomas,L.,K.P.Burnham,S.T.Buckland,S.T.Buckland,D.R.Anderson,K.P.Burnham,J.L.Laake,D.L.Borchers,

andL.Thomas.2004.Temporalinferencesfromdistancesamplingsurveys.Pages71-107OxfordUniversityPress

NewYork.

Thums,M.,S.D.Whiting,J.W.Reisser,K.L.Pendoley,C.B.Pattiaratchi,R.G.Harcourt,C.R.McMahon,andM.

G.Meekan.2013.Trackingseaturtlehatchlings-apilotstudyusingacoustictelemetry.JournalofExperimental

MarineBiologyandEcology440:156-163.

Trenkel,V.M.,L.Berger,S.Bourguignon,M.Doray,R.Fablet,J.Massé,V.r.Mazauric,C.Poncelet,G.Quemener,

andC.Scalabrin.2009.OverviewofrecentprogressinfisheriesacousticsmadebyIfremerwithexamplesfrom

theBayofBiscay.AquaticLivingResources22:433-445.

VanParijs, S.M., C.W.Clark, R. S. Sousa-Lima, S. E. Parks, S. Rankin,D.Risch, and I. vanOpzeeland. 2009.

Managementandresearchapplicationsofreal-timeandarchivalpassiveacousticsensorsovervaryingtemporal

andspatialscales.MarEcolProgSer395:21-36.

vanPolanenPetel,T.D.,J.M.Terhune,M.A.Hindell,andM.A.Giese.2006.Anassessmentoftheaudibilityof

soundfromhumantransportbybreedingWeddellseals(Leptonychotesweddellii).WildlifeResearch33:275-

291.

Vaneck, T. W., C. D. Rodriguez., M. C. Schmidt, and J. E. Manley. 1996. Automated bathymetry using an

autonomoussurfacecraft.Navigation43:407-419.

Verfuss,U.K.,D.Gillespie,J.Gordon,T.A.Marques,B.Miller,R.Plunkett,J.Theriault,D.J.Tollit,D.P.Zitterbart,

P. Hubert, and L. Thomas. Low Visibility Real-Time Monitoring Techniques Review. SMRUM-OGP2015-002,

providedtoIOGP.2016.

246


DATE:July2016


Verhoeven,G.J.,J.Loenders,F.Vermeulen,andR.Docter.2009.Helikiteaerialphotography-aversatilemeans

ofunmanned,radiocontrolled,low-altitudeaerialarchaeology.ArchaeologicalProspection16:125-138.

Vijay,S.TopicSeminaronAutonomousUnderwaterVehicles.P.E.S.CollegeofEngineering,2011.

Voegeli,F.A.,M.J.Smale,D.M.Webber,Y.Andrade,andR.K.O’Dor.2001.Ultrasonictelemetry,trackingand

automated monitoring technology for sharks. Pages 267-282 in The behavior and sensory biology of

elasmobranchfishes:ananthologyinmemoryofDonaldRichardNelson.Springer.

Vorontsova,M.Howtocarefullycountabelugaherd,includingeverybabybeluga.IFAW.2015.

Wall,C.C.,C.Lembke,andD.A.Mann.2012.Shelf-scalemappingofsoundproductionbyfishesintheeastern

GulfofMexico,usingautonomousglidertechnology.MarEcolProgSer449:55-64.

Wall,C.C.,C.Lembke,C.Hu,andD.A.Mann.2014.FishSoundProductioninthePresenceofHarmfulAlgal

BloomsintheEasternGulfofMexico.PLoSONE9:e114893.

Ward,J.A.,L.Thomas,S.Jarvis,N.DiMarzio,D.Moretti,T.A.Marques,C.Dunn,E.Hartvig,andP.Tyack.2012.

Passive acoustic density estimation of sperm whales in the Tongue of the Ocean, Bahamas.MarMam Sci

28:E444-E455.

Ward,J.A.,L.Thomas,S.Jarvis,N.DiMarzio,D.Moretti,T.A.Marques,C.Dunn,E.Hartvig,andP.Tyack.2012.

Passive acoustic density estimation of sperm whales in the Tongue of the Ocean, Bahamas.MarMam Sci

28:E444-E455.

Watts,A.C.,J.H.Perry,S.E.Smith,M.A.Burgess,B.E.Wilkinson,Z.Szantoi,P.G.Ifju,andH.F.Percival.2010.

SmallUnmannedAircraftSystemsforLow-AltitudeAerialSurveys.TheJournalofWildlifeManagement74:1614-

1619.

Webb,D.C.,P.J.Simonetti,andC.P.Jones.2001.SLOCUM:Anunderwatergliderpropelledbyenvironmental

energy.IeeeJournalofOceanicEngineering26:447-452.

Weber,N.,S.B.Weber,B.J.Godley,J.Ellick,M.Witt,andA.C.Broderick.2013.Telemetryasatoolforimproving

estimatesofmarineturtleabundance.BiologicalConservation167:90-96.

Weissenberger, J.,M.Blees, J.Christensen,K.Hartin,D. Ireland,andD.P. Zitterbart.Monitoring formarine

mammalsinAlaskausinga360°infraredcamerasystem.2011.

Williams,S.B.,O.Pizarro,J.M.Webster,R.J.Beaman,I.Mahon,M.Johnson,ÉRoberson,andT.C.Bridge.2010.

Autonomousunderwatervehicle-assistedsurveyingofdrowned reefson theshelfedgeof theGreatBarrier

Reef,Australia.JournalofFieldRobotics27:675-697.

Wolking,C. andA.Kleiner.AdvancedDevelopmentofAutonomousSemi-SubmersibleVehicles forMaritime

Applications.ReportbyC&CTechnologies,Inc.2011.

247


DATE:July2016


Wood,M.andR.Mahr.AutomatedDetectionofFishingVesselsusingSmartHydrophonesonanUnderwater

CoastalGlider.2016.

Wood, S. L. and C. E. Mierzwa. 2013. State of Technology in Autonomous Underwater Gliders. Marine

TechnologySocietyJournal47:84-96.

Wroblewski, J. S., B. G.Nolan, andG. A. Rose. 2000. Response of individual shoaling Atlantic cod to ocean

currentsonthenortheastNewfoundlandShelf.FisheriesResearch45:51-59.

Wroblewski, J. S.,W. L. Bailey, and K. A. Howse. 1994. Observations of adult Atlantic cod (Gadusmorhua)

overwintering in nearshorewaters of Trinity Bay,Newfoundland. Canadian Journal of Fisheries andAquatic

Sciences51:142-150.

Wynn,R.B.,B.J.Bett,A.J.Evans,G.Griffiths,V.A.I.Huvenne,A.R.Jones,M.R.Palmer,D.Dove,J.A.Howe,

andT.J.Boyd.InvestigatingthefeasibilityofutilizingAUVandGlidertechnologyformappingandmonitoringof

theUKMPAnetwork.FinalreportforDefraprojectMB0118.2013.NationalOceanographyCentre.

Wynn,R.B.,B.J.Bett,A.J.Evans,G.Griffiths,V.A.I.Huvenne,A.R.Jones,M.R.Palmer,D.Dove,J.A.Howe,

andT.J.Boyd.InvestigatingthefeasibilityofutilizingAUVandGlidertechnologyformappingandmonitoringof

theUKMPAnetwork.FinalreportforDefraprojectMB0118.2013.NationalOceanographyCentre.

Zhang,F.,D.M.Fratantoni,D.A.Paley,J.M.Lund,andN.E.Leonard.2007.Controlofcoordinatedpatternsfor

oceansampling.InternationalJournalofControl80:1186-1199.

Zitterbart,D. P., L. Kindermann, E. Burkhardt, andO. Boebel. 2013.Automatic round-the-clock detectionof

whalesformitigationfromunderwaternoiseimpacts.PLOS8:1-6.

Documents

LITERATURE REVIEW: UNDERSTANDING THE CURRENT STATE …€¦ · literature review: understanding the current state of autonomous technologies to improve ... report code: smruc-ogp-2015-015