AID AL EIEEIIEIIEIIIE - dtic.mil 677 experimental evaluation of simulator-based training for i/i marine pilots(u) national maritime research center kings point nv computer aid t j

AD-Ai59 677 EXPERIMENTAL EVALUATION OF SIMULATOR-BASED TRAINING FOR i/i

MARINE PILOTS(U) NATIONAL MARITIME RESEARCH CENTERKINGS POINT NV COMPUTER AID T J HRMMELL ET AL

UNCLASSIFIED MAR 85 CAROF-58-83i8-03 USCG-D-26-85 F/G 5/9 NLEEEEE/IEEEIE/EIEEIIEIIEIIIEIIIIIIIIIIIIII~flfl.EEllEEllEliEIlIllllIIhE~lllh

IIIIIIIIIIIIIIfllfll.

Lm

.25 11101 LA UEJ..6.

MICROCOPY RESOLUTION TEST CHARTN4ATIONAL. BUREAU OF STANDARDS-1963-A

REPORT NO. CG-D-26-85 CAORF-50-8318-03

TECHNICAL REPORT

EXPERIMENTAL EVALUATIONSOF SIMULATOR-BASED TRAINING

FOR MARINE PILOTS

' DTICwMarch 1985 ELECTE

ECTE

U.S. DEPARTMENT OF TRANSPORTATION U.S. DEPARTMENT OF TRANSPORTATIONMARITIME ADMINISTRATION UNITED STATES COAST GUARDOffice of Shipbuilding, Operations, and Research Office of Research and DevelopmentWashington, D.C. 20590 Washington, D.C. 20593

ioWts be. Upi

s .--=== 85 09 30 030.. . .. ' .,,... .,:,...:., , , ,,-p... .. .. ,, ... -....,......,:...-. ...... ....,.. .... . ..-2 .*. ...-..;:..... ...:...... . .

Accession For

NTIS GRA&IDTIC TABUnannounced ElJustification

13Distribution/Availability Codes

4 ;Avail and/or 0Dist Special

LEGAL NOTICE

This report was prepared as an account of government-sponsored work. Neither theUnited States, nor the Maritime Administration, nor any person (A) Makes anywarranty or representation, expressed or implied, with respect to the accuracy,completeness, or usefulness of the information contained in this report, or thatthe use of any information, apparatus, method, or process disclosed in this reportmay not infringe privately owned rights; or (B) Assumes any liabilities withrespect to use of or for damages resulting from the use of any information,apparatus, method, or process disclosed in this report. As used in the above,"persons acting on behalf of the Maritime Administration" includes any employee orcontractor of the Maritime Administration to the extent that such employees orcontractor prepares, handles, or distributes, or provides access to anyinformation pursuant to his employment or contract with the MaritimeAdministration.

LEGAL NOTICE

This document is disseminated under the sponsorship of the Department ofTransportation in the interest of information exchange. The United StatesGovernment assumes no liability for its contents or use thereof.

The United States Government does not endorse products or manufacturers. Trade ormanufacturer's names appear herein solely because they are considered essential tothe object of this report.

The contents of this report do not necessarily reflect the official view or policyof the U.S. Coast Guard and do not constitute a standard, specification, orregulation.

.A IJ6 77BIBLIOGRAPHIC DATA I. Rpeon No. 2. 3. Reipmt' Accemion No.

4. Title and Suttle a. Repor De

EXPERIMENTAL EVALUATION OF SIMULATOR-BASED TRAINING FOR March 1985MARINE PILOTS ,.

7. Authorisl G. Performins Orloniation Reort No.

T.J. Hammell, J.W. Gynther, and V.M. Pittsley CAORF-50-8318-039. Peferming Organiztion Name nl Addim 10. Projuet/Taek/Work Unit No.Computer Aided Operations Research FacilityNational Maritime Research CenterKings Point, New York 11024

12. Sponsoring Ormnlation Neme and Addrom 13. Type of Report & PeriodCovered

U.S. Department of Transportation U.S. Department of Transportation CAORF ReseRrch StudyMaritime Administration U.S. Coast Guard -Office of Shipbuilding. Operations, and Research Office of Research and Development 14.Washington, D.C. 20590 Washington, D.C. 20593

15. SuPples entery Notes

Project Managers: Dr. John Gardenier, USCG; Joseph Puglisi, MARADProject Monitor: LCDR George Naccara, USCG

16. Absuracs

"Ship bridge/shiphandling simulators are becoming increasingly recognized as a poten-tially beneficial training medium for attaining a number of selected deck officerskills. The utilization of such simulators for marine pilot training began severalyears ago in Europe. This report describes the conduct and evaluation of a prototypesimulator-based training program administered to twelve (12) U.S. pilots from several

*different pilot associations. The ship bridge/shiphandling simulator located at theComputer Aided Operations Research Facility (CAORF) was employed during the research.The prototype training program was designed to evaluate (a) the potential effective-

*ness of simulator-based training for pilots, (b) several specific skill areas forwhich simulator-based training appears appropriate and (c) the impact of pilot experi-ence on the benefits to be derived from such training.IC iod ;

17. Key Words Ol DomAment Anayalis. 17e. Oewipto. .

Ship bridge/shiphandling simulatorsSimulator-based trainingMarine pilot training-Rate-of-turn indicator,,Doppler speed log,RACONS,EmergeAcy shlphandling, .-

17b lndtifioulOpon-Endod Ien.

17c. COSATI FileWGreu

18. Ave lWdty Statement 19.3seurft Clmftihenm4Tla 21. Ne06 PeeApproved for Rolel Repen) UNCLASSIFIED 83NTIS -m O I 3.Pr

Sprigfildfirgiis ~ olUNCLASSIFIED

V, u,-¢emni Two ~

PREFACE

Any research effort of this magnitude - Jacobsen Pilot Service, Longcannot be successfully accomplished Beach, CAwithout the cooperation and assistance - Maritime Research Instituteof many people. The authors would Netherlandslike to express their thanks to all - Northeast Pilot Association,individuals and organizations who con- Newport, RItributed so much to the success of - Pilot Association for the Baythis project. In particular, we would and River Delaware, Philadel-like to especially acknowledge the phia, PAfollowing individuals and organiza- - Rotterdam Europort Pilots,tions for their contributions: Netherlands

- San Francisco Bar Pilots, SanFrancisco, CA

* Captain W. Lascelle (United New - Tampa Bay Pilots, Tampa, FLYork New Jersey Sandy Hook Pilot - United New York New Jersey SandyAssociation) and Captain J. Bradley Hook Pilot Association, Saint(Pilot Association for the Bay and George, Staten Island, NYRiver Delaware) for their assis-tance, particularly during the e The following organizations fordesign and conduct of the experi- providing individuals who partici-mental program at CAORF. pated in the experimental program

on the CAORF simulator:* The following organizations for the

cooperation, hospitality, and in- - Charleston (S.C.) Branch Pilotsight provided by their representa- Associationtives to members of the project - Corpus Christi Pilotsteam during visits, discussions, - Crescent River Port Pilotsand in several cases, vessel tran- - Jacobsen Pilot Servicesits: - Pilot Association for the Bay

and River Delaware- Hochschule fur Nautik, Bremen, - San Francisco Bar Pilots

West Germany - Tampa Bay Pilots- Brotherhood of German Pilots, - United New York New Jersey Sandy

West Germany Hook Pilot Association- Charleston Branch Pilot Associa-

tion, Charleston, SC * Members of the CAORF staff, parti-- Corpus Christi Pilots, Corpus cularly Captain H. 0. Travis, who

Christi, TX worked diligently to prepare the- Crescent River Port Pilots, New many aspects of the simulation for

Orleans, LA the conduct of the experimental- Exxon U.S.A. training program.- Hamburg Polytechnic Institute,

hamburg, West Germany *Captain C. DeBoer (Rotterdam

- Hosto PiotsHouton TXPilos) nd is taffwhoproide

_L .

the project team with a wealth of * The Government's project managers,information on pilot simulator Dr. J. Gardenier (U.S. Coast Guard)training, including the opportunity and Mr. J. Puglisi (Maritime Admin-to observe first-hand the simulator istration), and the U.S. Coasttraining of Dutch pilots. Guard Project Monitor LCDR G.

Nacarra for their guidance, assis-* Captain G. Zade and his staff at tance and timely contributions to

the Hochschule Fur Nautik who pro- the success of the project.vided the project with invaluableinsight relating to their experi- Finally, it should be noted that theences in utilizing the ship bridge/ support and participation by any ofshiphandling simulator for the the individuals or organizations citedtraining of German pilots, above does not necessarily imply their

agreement with all aspects of thise Captain C. Sander and Captain J. research project or this report. This

Froese at the Hamburg Polytechnic report represents the author's find-Institute who shared the benefits ings and recommendations, which may orof their planning as they prepared may not reflect the views of theseto commence simulator-based train- individuals and organizations.ing for pilots on their new ShipOperation and Simulation Facility(SUSAN).

Q d

* ] V

TABLE OF CONTENTS

Paragraph Page

CHAPTER 1 - INTRODUCTION

1.1 Background ............................................ 11.2 Project Goal .................................................. 21.3 Experimental Objectives........................................ 3

CHAPTER 2 - METHODOLOGY

2.1 Project Approach .................................. 52.2 Geographic Data Base .............................. 72.3 Experimental Variables ................................... 92.4 Experimental Design .............................. 102.5 P il1ot Part ici pants ................................ 102.6 Training Program .................................... 112.7 Experimental Procedures ................................ 142.8 Ownship Characteristics........................................ 142.9 Test Exercises........................................ 152.10 Performance Measures ................................... 16

CHAPTER 3 - RESULTS

3.1 Advanced Instrumentation - Module I................ ... 213.2 Emergency Shiphandling -Module II ......................... 26

CHAPTER 4 - DISCUSSION

4.1 Training Effectiveness .................................. 314.2 Areas of Pilot Training .......................... 314.3 Level of Experience .................................... 40

CHAPTERS5- CONCLUSIONS 57

CHAPTER 6 - RECOMMENDATIONS 61

APPENDIX A - DEBRIEFING QUESTIONNAIRE 65

B IBL IOGRAPHY 75

LIST OF ILLUSTRATIONS

Figure Title Page

1 Approach to Lotsedorf and Ball River .......................... 82 Experimental Design Summary ................................... 113 Training Schedule ............................................. 134 Segmentation of Test Exercise Turn for Data Analysis .......... 175 Module I Training Effectiveness -- Sum of Distance off 1 nm

Radius ...................................................... 336 Module I Training Effectiveness -- Mean Sum of Differences

Between Actual and Desired Rate-of-Turn ..................... 337 Module I Training Effectiveness -- Minimum Distance to Channel

Boundary .................................................... 348 Module I Training Effectiveness -- Number of Course Orders ... 349 Module I Training Effectiveness -- Number of Rudder Orders .... 3F10 Module II Training Effectiveness -- Minimum Distance to Channel

Boundary ................................................... 3711 Module II Training Effectiveness Number of Rudder Orders ... 3712 Module II Training Effectiveness -- Number of Course Orders ... 3813 Module II Training Effectiveness -- Number of Engine Orders ... 3814 Level of Experience, Module I -- Sum of Differences off 1.0

nm Radius .................................................. 4115 Level of Experience, Module I -- Sum of Differences Between

Actual and Desired Rate-of-Turn ............................. 4116 Level of Experience, Module I -- Minimum Distance to Channel

Boundary ................................................... 4217 Level of Experience, Module II -- Minimum Distance to Channel

Boundary .................................................... 4218 Level of Experience, Module II -- Number of Rudder Orders ... 4319 Level of Experience, Module II -- Number of Engine Orders ... 43

iv

ty A .6..

41

LIST OF TABLES

Table Title Page

1 Training Objectives for Prototype Simulator Training Programfor Marine Pilots ........................................... 23

2 Course Correction (Degrees) for Cross-Channel Current ......... 193 Module I - Overall Pilot Pretraining Versus Posttraining

Comparison ................................................. 224 Module I - Limited Experience, Pretraining Versus Posttraining

Comparison .................................................. 245 Module I - Extensive Experience, Pretraining Versus Posttrain-

ing Comparison .............................................. 256 Module II - Overall Pilot Pretraining Versus Posttraining

Comparison .............................................. ... 277 Module II - Limited Experience Pretraining Versus Posttraining

Comparison ................................................. 288 Module II - Extensive Experience Pretraining Versus Posttrain-

ing Comparison .............................................. 299 Pilot Questionnaire Summary Data for Selected Questions ...... 47

-.

!-.-.4 ..

W'.

V,°V

, o

% * *q4* ~ * ~ -

i

p

EXECUTIVE SUMMARY

BACKGROUND used.) Two senior pilots helpeddesign and conduct the training pro-

The U.S. Coast Guard provides for gram. The twelve trainees were pilotspartial credit toward certain mariner of varying experience from the U.S.licenses, endorsements or renewals East, Gulf, and West coasts. Trainingbased on completing "approved" simula- in three skill areas: Advanced In-tor training. Previous studies in strumentation, Emergency Shiphandlingthis series have produced the guide- and Decisionmaking (in collision emer-lines for approval of Simulator train- gencies) was administered. Objectiveing for cadets and for senior mariners measures of shiphandling performancewho are ship's company. No such were taken before and after each typeguidelines existed for approving simu- of training. Pilot opinions werelator training for local pilots. In recorded using debriefing question-fact, there was substantial doubt naires. Both objective data and sub-among pilots as to whether current jective opinions were analyzed tosimulator technology could adequately reach the conclusions of this study.display the shiphandling precisionneeded for sound pilot training. CONCLUSIONS

THE STUDY 1. Some forms of simulator trainingare highly effective in improving

Dramatic recent advances in ship simu- pilot skills.lator capabilities, along with con-firmed success of pilot simulator 2. Experienced pilots, apprentices,training programs in Europe, led to and qualified pilots with limitedthis research project. Its objectives experience can all benefit fromwere: First, to establish whether simulator training.valid and and credible pilot simulatortraining was feasible and, second, if 3. Different types of training areit was credible, to formulate approval needed for pilots with differentguidelines for training courses to levels of experience; however,justify licensing credit. The student pilots can benefit by mixing dif-was jointed sponsored and funded by ferent experience levels withfn thethe U.S. Coast Guard and Maritime same class.Administration. The Maritime Adminis-tratlon's Computer-Aided Operations 4. The simulated port area in a train-Research Facility (CAORF) was the ing program does not have to be aresearch site. An experimental train- specific real port, but it shoulding program was developed specifically present challenges that are rele-to measure training benefits. (It was vant to each trainee's own port.not designed to optimize training,although systematic instructional sys- 5. Pilots require higher fidelitytem development principles were training simulation, especially

vii

K L .' " ] " " % " "" - ' w ' " "- - . ' " • " ' '" ,". ".

regarding the fine points of ship such training was of value. A pro-motions, than do ship's company gram to optimize training, rathermariners. than measurement, would offer more

problems and more practice on each6. The experimental training program problem than this experimental pro-

in this study featured: gram.

a. Small class size (4-5 pilots).9. Identifying specific types of

b. A mixture of senior and less training for pilots (such as Ad-experienced pilots in the same vanced Instrumentation and Emer-class. gency Shiphandling) is aided by a

structured research approach de-*c. A seminar format with the in- scribed in the body of the report.

structor acting as a seminarleader, rather than as a author-

*itarian teacher lecturing stu- RECOMMENDATIONS FOR SIMULATOR TRAIN-dents. ING POLICY AND FOR FURTHER RESEARCH

d. Coordinators who were highly This study did not require policy andexperienced, active pilots and future research recommnendations; how-who were also familiar with the ever, the following are useful indica-functioning of the specific tors of the interests of the studysimulators. team and the pilot trainees.

e. Trainees who were volunteers 1. The Government should encourage,(not forced by Government to but not require, simulator trainingtake the training), for pilots.

7. Within our, sample of twelve pilots, 2. Several pilot groups for whomthe great majority felt: training would be particularly

cost-effective should be identi-a. That simulator training could be fied. Demonstration training pro-

beneficial and effective, if grams should be cost-shared by theproperly designed and conducted. pilots and the Government.

b. That high simulator fidelity is 3. Additional research should be con-essential. ducted, specifically:

c. That each of the training pro- a. More complete identification ofgram features in 6. above is the types of simulatnr trainingimportant and beneficial. beneficial for pilots.

d. That, during the experiment, the b. Development of a high qualityamount of time spent on each generic port data base foraspect of training was too short training a wide range of pilot

Vand allowed too few practice skills.runs.

8.Tewsc. Refinement and validation of8.Tetraining in this study wsship motion models suitable fordesigned to measure whether any simulator training for pilots.

viii

t4aA i...& -

d. At-sea validation of the bene- RELATED INFORMATIONfits of simulator training.

The International Marine Simulator4. The U.S. Coast Guard is not plan- Forum (IMSF) a worldwide assembly of

ning to pursue any such lines of visual ship simulator facilities, is

research. undertaking an international survey ofpilot simulator training needs andcapabilities.

ix

.°

, ,

CHAPTER 1

INTRODUCTION

1.1 BACKGROUND Pilots, who as a group represent thehigh extreme of shiphandling skills,

The ship bridge/shiphandling simulator have availed themselves of simulator-has become an important tool for use based training only to a minor ex-in research and training within the tent. This is due to several reasons,maritime industry. It has been used such as the lack of experience workingsuccessfully to train all levels of with simulators, the level of simula-deck officers, from cadets through tion technology available, and themasters. In some cases, simulator- lack of definitive evidence demon-based training is a cost-effective strating the effectiveness of simula-alternative for achieving specific tor-based training for pilots. Thetraining objectives, while in other most notable uses of simulator-basedcases it enables training and practice training for pilots occur in Europe --in situations that cannot be adequate- the Rotterdam Europort pilots wholy addressed via other means (e.g., train in Wageningen, Netherlands; thedue to safety considerations, or lack simulation facility at the Hochschuleof training control). It is generally Fur Nautik in Bremen, Germany whichviewed as a means of augmenting exist- has trained German pilots; and theing training programs. A variety of Port Revel model ship basin inshipping companies, for example, have Grenoble, France which has been usedmade available simulator-based train- to train various groups of pilots,ing programs for their senior level including pilots from the Uniteddeck officers, renting time at the States. The Europort pilots have beencommercial schools world-wide. At the training on the simulator at Wagenin-other extreme, all deck cadets at the gen for about 10 years. After ini-U.S. Merchant Marine Academy now partially trying the simulator with someticipate in a shiphandling simulator- degree of reservation, the Dutch havebased training program, using the come to rely on simulator-based train-government-owned Computer Aided Opera- ing as an important part of pilot pre-tions Research Facility (CAORF). Sev- paration and refresher training. Eacheral deck officer labor organizations, Europort pilot, who is a highly expe-(e.g., Marine Engineers Beneficial rienced professional bringing largeAssociation - District Two; Interna- tankers in and out of Europort, under-tional Organization of Masters, Mates, goes refresher training on the simula-and Pilots) have developed their own tor every two years. Their trainingshiphandling simulator-based training focuses on the use of electronic aidssystem to accommodate members. Hence, to navigation and difficult large shipa variety of shiphandling training maneuvers in a narrow channel (e.g.,system exist world-wide to accommodate turning a VLCC around). Althougha growing demand. there are several notable uses of

simulator-based training by pilots,the effectiveness, or lack of

1

effectiveness, of this training media the effectiveness of simulator-basedfor United States pilots has not been training of maritime academy cadets,demonstrated. chief mates, and masters. A major

product of this research is a documentSeveral major issues have affected the which contains shiphandling simulatoruse of simulator-based training by training system design and applicationU.S. pilots. The most dominant of guidelines developed for use by seniorthese is the lack of detailed objec- deck officers. The guidelines identi-tive information regarding the poten- fy the major characteristics of ship-tial use of this training media by handling training systems, theirpilots, its capabilities and limita- capabilities and limitations, andtions, requisite simulation character- recommend specific levels of trainingistics and quality, and so on -- that system fidelity and quality foris, details concerning its training addressing various areas of deck offi-effectiveness for pilots. Based on cer skill. A simulator functionalprevious research by the U.S. Coast design specification and applicationGuard and U.S. Maritime Administra- guidelines were also developed fortion, it was expected that simulator- maritime academy cadet training. Thisbased training would be effective for document identifies cadet/third mateaddressing certain piloting skills skills amenable to simulator-basedsuch as emergency shiphandling. training, recommends a specific train-Furthermore, it was expected that the ing system design for achieving a seteffectiveness of simulator-based of cadet training objectives, and pos-training of certain piloting skills sible detailed guidance informationwould depend on particular pilots for flexibly integrating simulator-backgrounds (e.g., the amount of expe- based training into academy curricularience). The greatest potential value as desired.of the simulator-based training ofpilots was not as a replacement for 1.2 PROJECT GOALparts of current pilot training pro-grams, since these are believed to be The goal of CAORF's Maritime Piloteffective. Rather, the potential was Training Investigation is to assistto augment current training programs the responsible parties in determiningwhere desired by particular groups of the appropriate role of the shippilots. Hence, the necessity existed bridge/shiphandling simulator withinto develop objective information the training process of marinedetailing the potential effective use pilots. It was the purpose of theof simulator-based training by pilots, experiment reported herein to extendand to provide guidance information to the empirical investigation of simula-assist pilots in using this technology. tor-based training to pilots. The

techniques developed during theThe U.S. Coast Guard and U.S. Maritime earlier research in this program wereAdministration, in response to the tailored for this investigation. Thegrowing availability of the ship- primary focus of the research was onhandling simulator training techno- the effectiveness of simulator-basedlogy, initiated research to determine training of experienced pilots.the cost-effectiveness of such train- Secondary issues addressed were theing media. This research, which has effectiveness of such training forbeen conducted by the Computer Aided different areas of piloting skill; theOperations Research Facility (CAORF), effectiveness of such training as ahas included experiments investigating function of the level of pilot experi-

2

~ 7 .*.***. ~ ..-- -*- - -. ***J ~ .z. .

ence; and, the opinions of simulator- program for pilots, which was con-wise experienced pilots regarding the ducted on the CAORF simulator, were:potential of such training. Anexperiment, was conducted at CAORF to e Can a shiphandling simulator beinvestigate these issues. Its purpose used effectively to assist marinewas to generate objective information pilots in acquiring selected skills?addressing the effectiveness of pilottraining on shiphandling simulators. * What types of benefits shouldThis experiment is the subject of this pilots anticipate from simulator-report. Another major product of this based training in specific skilleffort was the development of guide- areas?lines addressing the design and use ofthe simulator-based shiphandling . How do these benefits from simula-training system by pilots (Gynther, tor-based training vary with theHammell and Pittsley, 1984). These experience level of the pilot?guidelines have been developed to beused by pilots when deciding whetheror not to use simulator-based train- These are major issues for which thising, when evaluating different avail- experiment could develop objectiveable training systems, and when devel- data. Although it was beyond theoping or using a simulator-based scope of this experiment to comprehen-training program. sively address each of these issues,

it was believed that they could be1.3 EXPERIMENTAL OBJECTIVES addressed to an acceptable level of

depth to provide guidance to the pilotThe specific issues that were investi- community regarding the use of simula-gated during the experimental training tors for training.

3

CHAPTER 2

METHODOLOGY

2.1 PROJECT APPROACH and allocated to the various trainingmedia normally available for pilot

The approach followed in this empiri- training -- classroom, at-sea, smallcal investigation was similar to that vessel, and simulator training. Thisin earlier investigations of this allocation was determined by identify-long-term program. It was based on a ing the media that appeared to be thesystematic approach to training, using most cost-effective means of achievinginvestigative techniques developed by each training objective. The resultthe Air Force and Navy (i.e., instruc- of this process was a subset of marinetional systems development process, pilot training objectives that ap-systems approach to training, training peared to be appropriate for accom-system analysis), and modified as ne- plishment via simulator-based train-cessary for application in the mani- ing. The skill areas identified fortime industry, and for specific appli- which simulator-based training iscation to the issues'of this investi- potentially a cost-effective mediumgation. A summary overview of the were:approach follows.

. Emergency Shiphandling (mainly ram-First, the tasks that pilots perform ming and grounding avoidance prob-at-sea were identified. This was lems)accomplished by the project team aftera review of the deck officer task * Shiphandlinglisting developed during the firstphase of the Training and Licensing e Vessel Characteristics (unfamiliarproject (Hanmmell, Williams, Grasso, vessels)and Evans, 1980), a review of thevideo tapes of piloting activities * Pilothouse Proceduresoriginally obtained during the Huffnerstudies (Huffner, 1978), visits to * Advanced Instrumentationseveral pilot associations, ship ridesin several pilotage areas, and discus- * Restricted Waters Navigationsions with numerous pilots. Thesetasks were then reviewed by one of the e Rules of the Roadproject's pilot consultants and appro-priate alterations made to develop a * Vessel-to-Vessel Communicationsfinal set.

* Restricted Waters DecisionmakingNext, the tasks were analyzed into (mainly reaction to collision situ-skill and knowledge elements, which ations)

.4 were then transposed into a listing oftraining objectives. These training This listing was then further scruti-objectives were then further evaluated nized to identify representative skill

5

areas for evaluation during this CAORF Weak Interestexperiment. At this point in the ana-lysis, the following issues were con- e Area familiarization (not involvingsidered: (a) the simulator capabili- changes to the port)ties that would normally be availableat a commercial training facility, (b) Although the training areas of inter-the European simulator-based pilot est are grouped and labeled different-training experiences, and (c) the ly, the interests are quite common.causal factors identified in various Local area familiarization is of some-published accident reports which what weak interest mainly becauseappear to indicate pilot skill areas pilots feel that existing area famil-

- ~ that may benefit from additional iarization training programs are quitetraining. adequate. Not that some degree of

familiarization cannot be accomplishedThis list was validated by comparison on a simulator, but rather that simu-to independently sampled European lator training is not fully adequateopinions. Experienced German pilots, to the task and that at-sea orienta-polled informally for their interest tion is adequate. The master-pilotin simulator oriented training, relationship has been previously iden-expressed various degrees of interest tified (Gynther, Haminell, Grasso andin the following training topics: Pittsley, 1982) as a pre-eminently

critical shiphandling requirement for%Greatest Interest restricted waters navigation. To

date, however, simulator training hase Restricted waters maneuvers of con- only evolved to address masters and

cern to the pilots themselves pilots separately.

* Emergency situations (including A prototype training program was de-equipment failures)* veloped for experimental purposes and

conducted via the simulator at CAORF.9 Communications ship to ship and This research, using the developed

ship to shore training program, had the dual role of(1) generating objective data with

Moderate Interest which to further assess and documentthe potential use of simulator-based

e Outside the simulator review of training of pilots and (2) providing abasic theory and analyses of speci- valuable base of data and experiencefic relevant casualties for development of the previously

noted consumers guide relating to thee Advanced instrumentation use of simulators for training pilots.

e Master-pilot relationships As previously noted, this experimentinvestigated three major issues: (1)

9 Vessel familiarization with special the effectiveness of simulator-basedvessels training of pilots; (2) the effective-

ness of such training for three dif-* Familiarization with significant ferent areas of piloting skill; and

changes to port waterways (3) the effectiveness of such trainingfor two different levels of piloting

*Including what this study def ined as experience. Pilots from East, Gulf,decislonmaking and West Coast ports participated in

the experiment. Both test and train- most effective pilot training, it ising exercises were conducted on the necessary to consider training in asimulator during each of three four- generic-port data base due to costday training programs. Classroom ses- considerations.sions were interspersed with the simu-lator exercises as part of the train- The viability of pilot training in aing program. The effectiveness of the generic port data base was, therefore,simulator-based training was deter- a major factor in this experiment.mined on the basis of group perf or- The effectiveness of simulator-basedmance in certain controlled simulatortraining for pilots was investigatedbased shiphandling exercises that in a generic port data base developedimmediately preceded and followed seg- specifically for this experiment.ments of the training program. It was This geographic data base was a hypo-assumed that the performance of the thetical port specifically designed topilots on the simulator during both assist in the refinement of the skillsthe pretraining and posttraining exer- selected for each training module. Acises would be representative of their chart of this hypothetical port,performance during similar piloting Lotsedorf, is contained in Figure 1.operations at-sea. Observations and Both container and petroleum bulkcomments by the pilot coordinators and cargoes are handled in the port. Apilot participants were also factored container terminal is located atinto the analysis. Lotsedorf. The petroleum terminal is

located further up the Ball River. AnThis section addresses the major as- anchorage is located in Burton Bay forpects of methodology used during the vessels waiting for a berth at eitherpreparation and conduct of the experi- the cargo terminal or the oil termi-ment. The results of the experiment nal. The topography is barren andare documented and discussed in rocky. When approaching from seaward,Chapters 3 and 4. passage must be made via a winding

800-foot-wide channel through danger-2.2 GEOGRAPHIC DATA BASE ous shoal areas. The channel is

buoyed; however, ice and frequentIt has been a common belief that a storms may disperse these aids suchport-specific geographic data base that mariners should navigate withwould be necessary for the simulator- extreme caution. The beacons oun

'based training of pilots. This would Shea's Ledge, North Ledge, and Southrepresent a very expensive approach if Ledge also provide fixed frequencysuch training were to become a reality racon coverage.for a large segment of the pilot popu-lation. Furthermore, the simple cost- There are also several potential bene-effectiveness of simulator-based fits for conducting pilot training intraining would be seriously questioned generic data bases. First, pilots mayif it is necessary to develop port- be more receptive to generic data basespecific data bases for the smaller training because they can participatepilot associations (e.g., 15 to 20 in training without (a) focusing theirmembers). Hence, a more cost-effec- attention on the "inadequacies" of thetive alternative to the port-specific training device for simulating theirdata base is necessary. Although it specific port, or (b) perceiving thatis believed by many individuals, par- they are putting their credentials forticularly pilots, that port-specific pilotage in their port on the line ifdata bases are likely to yield the they should make a mistake. Second,

* C2

NORTH LEDGE

(A IFIXEDi WHITE) IRACON -D-) 1.5NM

NtI O..N

SMITH-Nil m3t, SHEA'S LEDGESM1T ANNIC.FV LIG~HT

POINT 'CieA CX~I CIS (FIXED WHITE)

II~ oNm SOUTH LEDGE RCN-INIS -M' LIGHT~*' ,'(FIXED WHITE)~# ~ b2 RACON --

.41

-' LOTSEDORF POINT

IN E

N23j.

BURTON JSAY I IC2 LOTSEDORF

N27 I c~qIIh LENCTH NDEPTH INOWEIDCHANEL.. M (E?) 11m) ILM IPT? COUfM

CK I 191111 OW 2.3 45 ne0

mMJUOEL -4 .1 -6 mema 1I"' mlw 4.0 me 1

CHAII ~ tt -2Ulw A l

Figure 1. Approach to Lotserdorf and Ball River

8

-1111 1 -111 5

the utilization of generic data bases skill area, advanced piloting instru-for pilot training would probably also mentation, involved refining a pilot'smean that a given simulator class ability to effectively employ some ofwould involve pilots from several geo- the more sophisticated electronicgraphic locales. A valuable cross- navigational equipment that arefertilization of expertise may take appearing more frequently on today'splace, particularly for experienced ships. Pilots, particularly experi-pilots, since it would potentially enced pilots, are well-versed with theallow the exchange of information con- equipment that they employ duringcerning a variety of successful pilot- their transits. However, severaling techniques. Third, for training independent sources have indicatedspecific skills, such as advanced that pilots may benefit significantlyinstrumentation, a simple generic data from additional training in utilizingbase may focu5 student attention various modern electronic equipments,better than a data base with specific such as stabilized radar, racons,port cues. doppler speed log, and rate-of-turn

indicator. Skills related to the use2.3 EXPERIMENTAL VARIABLES of advanced instrumentation have been

a major area of simulator-based train-The principle variable investigated in ing for pilots at the Wageningen (thethis experiment was the effectiveness Netherlands) and Bremen (Germany)

*of simulator-based training for training facilities.*pilots. This is a bottom-line issue

of concern to pilots, directly relat- The second skill area, emergency ship-ing to the cost/benefit ratio for handling has been generally recognizedpilot training using simulators as a as an area for potentially effectivesupplement to the existing pilot training on a simulator. Most pilotstraining programs. The training are receptive to this type of simula-

*effectiveness was addressed in terms tor-based training since it obviouslyof the improvement/decrement in cer- can not be trained effectively at seatain piloting skills that resulted due to the high risks involved. Thefrom participation in the training piloting skills associated with keep-program. ing a large vessel under control when

executing a crash stop in a narrowTwo secondary variables were investi- channel, or turning a large vesselgated in addition to the training around within a confined channel with-effectiveness -- pilot skill areas, out tugs, although infrequently used,and pilot level of experience. Each are well-recognized as important andof these secondary variables was desirable by many pilots. This typeaddressed in terms of its impact on of training on a simulator is alsothe effectiveness of simulator-based currently provided to European pilots.training. These secondary variablesare discussed below. The third and final pilot skill area

investigated was decisionmaking inPilot Skill Areas emergency or unusual restricted waters

situations. Discussions with many*Three pilot skill areas were examined pilots have indicated the importance

in this experiment: (a) advanced of being able to rapidly respond topiloting instrumentation, (b) emergen- common unforeseen difficult problems,cy shiphandling and (c) restricted such as a vessel unexpectedly depart-waters decisionmaking. The first ing from an anchorage and crossing

9

ownship's bow. These types of situa- eral unavailability. As a result oftions often involve the rapid detec- the reduced shipping in recent years,tion of a problem, the assessment of many pilot associations do not pre-alternative actions available, the sently have apprentices, nor do theyselection of a course of action, and anticipate accepting apprentices forthe effective implementation of that several years.action. The use of a simulator totrain this area of pilot skills 2.4 EXPERIMENTAL DESIGNappears especially suited to pilotswith a relatively limited amount of The experimental design for thisexperience, since it apparently takes experiment is summiarized in Figure 2.many years of experience to encounter A 1-week experimental training programthe range of potential problems and to was provided to three groups ofdevelop the skills necessary for early pilots, one group each week. Each ofdetection and handling of these situa- these groups were composed of fourtions. The simulator has potential exparienced pilots. Within each groupfor this area of training since it can there was a mix of limited experiencebe used to provide a pilot with a re- and extensive experience pilots. Eachpresentative set of problem situations 1-week training program was dividedin a relatively short time, which into three independent trainingotherwise might require many years of modules, with each module addressingexperience. Also, it should also be one of the pilot skill areas discussednoted that the risk of such training above. The objective data used forwould be substantially less than at training effectiveness evalt'ationssea.* were based on pre and postexercises

for each pilot, individually conductedPilot Experience before and after each training

module. Simulator difficulties pre-The effectiveness of simulator-based vented obtaining data for the thirdtraining is likely to depend on many training module within the experiment-factors. Pilot skill area is, per- al program. As a result, the objec-haps, the most relevant of these fac- tive data derived from shiphandlingtors. Another important factor may be performance on the simulator exercisesthe level of pilot experience. Simu- were available for training Modules Ilator-based training may be more and I I. Pilot observations and opin-effective for highly experienced ion data were collected for eachpilots regarding certain skill areas, module, including some data for theand perhaps more effective for pilots third training module. Chapters 3 andwith limited experience regarding 4 contain a detailed presentation andother skill areas. Two levels of discussion of the results obtainedpilot experience were examined regard- during this experimental program.I ing the effectiveness of simulator-based training: (a) limited experience 2.5 PILOT PARTICIPANTS(less than 5 years as a licensedpilot), and (b) extensive experience Twelve licensed pilots participated in(greater than 10 years as a licensed this experimental CAORF program.pilot); see Pilot Participants (Para These pilots were equally distributedgraph 2.5) for additional information from East Coast, Gulf Coast, and Westregarding the qualifications of parti- Coast ports (i.e., four pilots fromcipating pilots. Apprentice pilots each coast) . The following organiza-were not considered due to their gen- tions provided participants:

10

WEEK I

4 PILOTS.2 LIM EXP__ _ ___

*2 EXTN EXP

LIMITED PRE 5 5 5

WEEK 2 EXPERIENCE

4 IOSEXPERIMENTAL PIOS POST 5 5 5.2 LM EX -~ TRAINING

EXTENSIVE PRE 7 7 7EXPERIENCED

PI LOTSWEEKP3 POST 7 7 7

4 PILOTS

*1 LIM IEXP*3EXTN, EXP

r

Figure 2. Experimental Design Summary

* United New York New Jersey Sandy years experience. The average age ofHook Pilot Association these individuals was 39.8 years, with

their ages varying from 26 to 56# Pilot Association for the Bay and years. Ten of the twelve pilots were

River Delaware between 30 and 50 years old. None ofthe pilots had previously participated

e Charleston (S.C.) Branch Pilot As- in a training program on an electronicsociation ship bridge/shiphandling simulator,

although several had previously beene Tampa Bay Pilots to CAORF as test subjects during other

research experiments. One pilot had@ Cresent River Port Pilots (New participated in the simulator training

Orleans) program at Grenoble, France whichemploys physical ship models.

* Corpus Christi Pilots 2.6 TRAINING PROGRAM

e San Francisco Bar PilotsThe prototype simulator-based training

e Jacobsen Pilot Service (Long Beach) program for pilots that was conductedat CAORF as part of this experiment

The average experience across all the addressed only a subset of the pilotparticipants was 9.0 years as a li- skills that analysis identified ascensed pilot. The range of this potentially appropriate for develop-experience was from 1.5 years to 23.0 ment/improvement via simulator-basedyears. There were five pilots with training. Limited resources prohib-less than 5.0 years experience and ited the investigation of additionalseven pilots with greater than 10.0 skills. The treader should note that

1,11

" " " ', ",'.,'."'. ".,','"', ', ',.',''." '.'" "\ '- '. "- ", " '"v, -"'"'- ,' i " - "- '"¢ ":1

this experimental training program was TABLE 1. TRAINING OBJECTIVES FORdeveloped specifically for this ex- PROTOTYPE SIMULATOR TRAINING PROGRAMperiment, and thus is not necessarily FOR MARINE PILOTSa training program that should be of-fered to pilots at a commercial train- Training Module I: Advanced Pilotinging facility. As an experimental Instrumentationtraining program it was designed tocover several skill areas in order to * The trainee shall be able to effec-provide an indication of the possible tively execute a constant radiusbenefits from simulator-based train- turn under various operational con-ing. It was not designed to compre- ditions while employing a rate-of-hensively address these skills to the turn indicator, a doppler speedlevel of refinement that might be log, and racons.accomplished at a simulator facilitydedicated to training. Training Module II: Emergency Ship-

handling SkillsThe experimental training program wasa 1-week program designed for a class *The trainee shall be able to safelyof four pilots. It was composed of turn an 80,000 dwt tanker aroundthree training modules, which were within a confined channel underbased on the three skill areas identi- various operational conditions.fied and discussed in Paragraph 2.3.These three modules were (a) Advanced Training Module III: DecisionmakingPiloting Instrumentation, (b) Emergen- in Unusual or Emergency Restrictedcy Shiphandling, and (c) Restricted Waters SituationsWaters Decisionmaking. The specifictraining objectives developed for each e The trainee shall be able to effec-module are presented in Table 1. tively respond to an unanticipated

vessel departing from an anchorageThe structure of the training program or berth within a confined channelinvolved a combination of classroom under various operationaland simulator training. It is out- conditions.lined in Figure 3. The training stra-tegy used in the program was a modifi-cation of seminar, role playing, and the classroom period. During thiscase study techniques, which were con- group exercise period, the traineessidered appropriate for skilled pro- were rotated between the positions offessionals focusing on cognitive pilot, helmsman, and observer.tasks. Each training module commencedwith the trainees individually han- The training program was coordinateddling the ownship in a controlled by two highly experienced and re-simulator exercise. This exercise was spected pilots currently active inthen discussed during the classroom different pilot associations. Each ofsession as a case study example of a these coordinators are articulate in-situation requiring particular skills, dividuals as regards piloting andwhich would be addressed during the piloting techniques. They possessedtraining module. After the classroom previous experience with shiphandlingsession, the pilots participated in a simulators, and were versed with manyseries of exercises on the simulator. of the research protocols employed atThey applied and practiced the princi- CAORF. As coordinators they were the

ples and techniques discussed during primary interface with the pilot par-

12 I

TRAINING MODULE NO. 1: TRAINING MODULE NO. 2: TRAINING MODULE NO. 3:

M T W _T F

0800 0800 . .... 0800 - --- 0800

C'LASSq

...... 0900

'hi"SIM EX:INDI~t "G)dUPStim EX: SIM EX.

1000GRU

1100SIM E X

* . ~1100 11010.....

::jGROUPe.-CLS-SIM EX,:

1200 ..... %120 12001

LUNCH LUNCH *:INDIV. LUNCH:SIM EX

1300 1300 1300

1400 1400

CLA S1500 moNV

SIM EX SM EXINTRO

160~ 1600SIM EX

ASIM FAM X1700 .... DEBRIEF

DINNER l300 1800 1800 1800o

- LEGEND:

SIONS SIM FAM - SIMULATOR FAMILIARIZATION

INDIV SIM EX - INDIVIDUAL SIMULATOR EXERCISESGROUP SIM EX - GROUP SIMULATOR EXERCISES

21001

Figure 3. Training Schedule

13

ticipants, providing tcnalinfor- each test the exercise, he was in-mation, overseeing the various train- structed to refrain from discussinging activities, and directing the the exercise with other pilotstraining program. They also assisted until the next class session. A

*during the development of the experi- variety of pertinent maritime filmsment and the training program. were available for viewing by the

pilots when they were not on the2.7 EXPERIMENTAL PROCEDURES simulator during the testing peri-

ods.Extreme care was exercised in conduct-ing the training program to minimize d. During the training periods, theuncontrolled aspects of the experiment pilots were rotated through thewhich can result from variations in position of pilot and helmsman.the selection, training, and debrief- During Modules I and II thoseing of subjects. The following proce- pilots not participating in thedures were followed: exercise were located on the bridge

* in an unobtrusive location with thea. Upon initial contact, via his re- coordinator. During Module III,

spective pilot association, the the most advantageous position forpilot participant was appraised of the off-watch pilots and the in-the experimental nature of the structor was determined to be attraining program. He was told that the human factors monitoring sta-

iis anticipated he would receive tion, where they could anticipate,beeiiltraining; however, his observe, and discuss exercise

primary purpose was to assist the events without disturbing thoseCAORF staff in the evaluation of pilots handling the simulated yes-this type of simulator-based train- sel.ing for pilots.

e. Upon completion of the training,b. Upon arrival at CAORF, each pilot each pilot was individually de-

was provided with additional infor- briefed by a member of the researchmation during the introductory ses- staff, who completed the debriefingsions concerning the purpose of the questionnaire contained in AppendixPilot Training Investigation, the A.objectives of this particular ex-perimental training program, andhis role during the coming week. 2.8 OWNSHIP CHARACTERISTICSHe was also provided with an appro-priate familiarization of the CAORF The following are characteristics ofsimulator including an opportunity the simulated ship that was used dur-to handle the simulated vessel in ing the experimental training programthe generic data base used during at CAORF:

C. the program.. Type: 80,000 dwt tanker

c. During testing periods, each pilotwas on the simulator bridge alone * Length overall (LOA): 800 feetexcept for a helmsman, who was a

*member of the CAORF staff. The e Length between perpendicularspilot was provided with appropriate (LBP): 763 feetinstructions prior to the commnence-

*ment of each test exercise. After * Location of pilothouse: Aft

* 14

* Beam: 125 feet *Test Exercise 1 - The 80,000 dwttanker (fully loaded) was posi-

* Draft (full load): tioned inbound on the centerline of- Fwd - 39.9 feet Channel C at position A in Figure- Aft - 39.9 feet 1. Initial speed was 8 knots. It

was a clear day with unrestricted* Propulsion: visibility. A 1.0 knot current was

- Slow speed diesel setting towards 0900. Buoys C8- 23,000 horsepower and C10 were missing as a result of- Single screw a recent storm. The trainee was- Direct pilothouse control directed to safely negotiate the

1.0 nm radius turn between Channel2. TS EECIE alal Channel D utilizing the

2.9 ESTEXECISS aailbleaids to navigation. Thistest exercise was designed to pro-

Four test exercises were developed for vide pilot performance data priorthe experimental training program. to Training Module I.The purpose of these test exerciseswas to generate the objective pilot e Test Exercise 2 - The 80,000 dwtperformance data used to assess the tanker (fully loaded) was posi-gain/loss in performance that resulted tioned again inbound on the center-from each training module. Test exer- line of Channel C at position A.cise 1 was designed to evaluate Initial speed was 8 knots. Visi-trainee performance prior to Training bility was unrestricted, and a 1.0Module I. Test exercise 2 was de- knot current was again settingsigned to provide both (a) posttrain- towards 0900. As in test exer-ing performance data for Module 1, and cise 1, buoys C8 and CIO were miss-(b) pretraining performance data for ing. The trainee was directed toModule II. Similarly, test exercise 3 safely negotiate the 1.0 nm radiuswas designed to provide (a) posttrain- turn between Channel C and Channeling data for Module 11, and (b) pre- D utilizing the available aids totraining data for Module III. Final- navigation. Upon completion of thely, test exercise 4 was designed to turn and entry in Channel D, theprovide posttraining performance data pilot was told that he should turnfor Module III. Due to simulator dif- the ship around in Channel D with-ficulties, which affected the training out tugs since there had been anprogram schedule, test exercise 4 was accident which was obstructing thenot accomplished. As a result, the channel at buoy N19. The 1.0 nmobjective performance data required to radius turn was designed to provideevaluate the training effectiveness of pilot performance data after ModuleModule III was not available. The I, while the turn around was de-subjective evaluation accomplished at signed to provide performance datathe end of the program, however, was prior to Module II.conducted for Module III using twogroups of pilots. The following is a e Test Exercise 3 - The 80,000 dwtbrief summary of each of the three tanker (fully loaded) was posi-test exercises that were accomplished tioned outbound on the centerlineand utilized as the basis for the ob- of Channel D at position B in Fi- Vjective analysis described in Chapters gure 1. Initial speed was 83 and 4. knots. Visibility was unrestricted

15

and a 1.0 knot current was setting transformed into performance measurestoward 2700. The trainee was relevant to the particular trainingtold that he should turn the ship objectives, and then statisticallyaround in Channel D without tugs analyzed to evaluate training effec-since there has been an accident tiventss.which is obstructing the channel atbuoy N7. Several performance measures were

identified for the types of pilotOnce the ship is turned around, the skills involved in Modules I and II.pilot was directed to return to his These performance measures were or-berth at the oil terminal. As own- ganized into three categories for eachship proceeded inbound the visibil- module, based on their relevance toity was gradually reduced to ap- the training objectives of the respec-proximately 0.5 nm. As ownship tive module: (a) primary performanceapproached the container terminal, measure, (b) secondary performancea containership unexpectedly backed measures, and (c) third-order perfor-into the channel less than 1.0 nm mance measures. The primary perfor-ahead. The containership desired mance measure is the single measureto proceed outbound. The earlier most applicable to the particularturn around within Channel D was training objectives, pilot skills, anddesigned to provide post-Module II relevant aspects of shiphandling. Itperformance data, while the unanti- represents the bottom-line with regardcipated backing of the container- to each respective training programship was designed to provide per- module. Previous research of thisformance data regarding pilot de- type in the maritime field has shown,cisionmaking skills prior to Module however, that for various reasons sin-Ill. gle performance measures may not be

insensitive to differences in perfor-2.10 PERFORMANCE MEASURES mance. Hence, multiple performance

measures are necessary to understandTwo major sources of data were used in the processes involved. Furthermore,the experiment: (1) objective perfor- these additional performance measuresmance data collected during the test address other relevant, althoughscenarios; and (2) subjective observa- secondary and third-order, aspects oftions and opinions of the pilot parti- performance and behavior. They assistcipants, collected during the debrief- in analyzing and interpreting the

, ing at the end of the training pro- experimental effects by providing datagram. The subjective data are com- regarding associated facets of perfor-prised of straightforward pilot opin- mance.ions concerning the specific issuesaddressed. Their analysis is in Para- A description of the performance mea-graph 4.4 in terms of the proportion sures used with each module follows.of various opinions expressed regard-ing each issue/question. The objec- Module I - Advanced Instrumentationtive data, on the other hand, werecomprised of various ownship and situ- The focus of this training programation parameters, some of which were module was the effective utilizationautomatically recorded by the simula- of advanced instrumentation, namelytor/computer and some which were manu- the rate-of-turn indicator, theally recorded by an observer. In both doppler speed log, and the availablecases, these objective data were racons, to execute a constant radius

16

,*.'.' .. * * - ., -. .' "'. , * " *-*'.', . - .". .." .." *.*.., .. . -." , -. *.." .. . 2- :. -." -. V -" .-. , * -.. ,". , .,

turn. The ability to achieve a con- placed at the end of the com-stant radius turn using these instru- plete turn. Ideally, the vesselments provides a set of piloting should have maintained a preciseskills that allows considerable vessel 1.0 nm radius from the raconcontrol during the navigation of throughout the turn. This con-turns; as such, it represents one stant radius was the objectiveeffective piloting technique, to aug- of training in Module I. Thement other available techniques. summation of this ownship posi-

tion deviation at each of the* Primary Performance Measure data lines provides a relevant

measure of the degree to which- Sum of Distances Off 1.0 nm Ra- the pilot was able to maintain

dius: This performance measure the constant radius.ithe sum of the absolute dif-ference in feet between the • Secondary Performance Measureideal 1.0 nm radius and theactual position of the vessel's - Sum of Differences Between Actu-center of gravity, at four dif- al and Desired Rate-of-Turn:ferent data lines through the This measure is similar to theturn. These data lines were previous performance measure inspaced at intervals of 1/4 of that it is calculated at speci-the turn, DL2 through DL5 in fic data lines in the turn. ToFigure 4. The first data line achieve the constant 1.0 nm ra-(DL2) was located at the end of dius turn, the vessel had tothe first one-fourth of the maintain a rate-of-turn of ap-turn, and the final data line proximately 0.11 deg/sec. This

DL4 '.DL3D

ExerciseDL5 Initiation

w

Racon

Figure 4. Segmentation of Test Exercise Turn for Data Analysis

17

value, which has both a theore- channel without tugs. The specifictical and empirical basis, as- test exercise involved turning ansumes an average speed over the 80,000 dwt tanker around within a 1500ground of approximately 7 knots ft channel without tugs, and with aas the ship proceeds through the 1-knot current on the starboard bow.turn. (Note: the theoretical Although some pilots may not considerbasis was provided to the this exercise as emergency ship-trainees as part of the class- handling training, most pilots agreedroom training.) Hence, this that it would not normally be at-measure addresses the differ- tempted except under emergency condi-ences between the ideal rate-oftions. The Dutch pilots during theirturn and the actual rate-of-turn simulator training at Wageningen de-achieved. The data lines at veloped an effective technique forwhich the rates-of-turn were safely accomplishing this maneuver.measured were identical to those This technique was the focus duringemployed for the previous mea- the emergency shiphandling module ofsure, with the exception that the CAORF experimental training pro-the final data line at the end gram. The technique involves balanc-of the turn was not used (DL5). ing the cross-channel component of theThis data line was omitted since ship's speed against the cross-channelthe vessel, having completed its component of the current, according toconstant radius turn, should Table 2, as the vessel slows. Thehave a zero rate of turn at that pilot applies a course correction topoint, head up into the current based on his

speed through the water at appropriate, Third-Order Performance Measures intervals of time. For the test exer-

cise, this method is extremely helpful- Number of Rudder Orders -- the for the first 30 degrees of course

number of rudder orders issued change (i.e., nearly heading into theduring the exercise current).

- Number of Course Orders -- the @ Primary Performance Measurenumber of course orders issuedduring the exercise - Channel Excursions: This mea-

sure is the number of pilotsMinimum Distance to Channel that had any part of their yes-Boundary --the closest point sel extend beyond the channelownship came to the channel boundaries. As such, it isboundary during the exercise clearly a bottom-line perfor-CE nnmance measure for the situation

- Channel Excursion -- the number addressed in Module II. Channelof pilots that had any part of excursions were determinedtheir vessel extend outside the graphically by viewing geograph-channel boundaries during the ic plots containing the hullexercise outline of ownship throughout

the exercise.Module II - Emergency Shiphandling

, * Secondary Performance MeasureThis module emphasized the refinement

of skills involved when turning a - Minimum Distance to Channellarge tanker around within a confined Boundary: This measure repre-

}-.

18

L,. N( LA-,;~ ,- & r~ om

TABLE 2 O CORRECTION (DEGREES) FOR CROSS-CHANNEL CURRENT

COMPONENT CURRENT . CHANNEL

(MILES/HOUR)

1/2 1 14 2 2,2 3

.. 9 3 6 9 12 is 18

< 8 3', 7 10V 14 17a 21

7 4 8 12 16 20/a 25

LU 6 5 10 15 20 25 30

T 5 6 12 18 24 30 36S 4 7', 15 22', 30 372 40a: 3 10 20 30 42 57 90

-2/ 12 24 36 53 90

W 2 15 30 40 90wU - - -

u 1/2 20 42 90

1 30 901

0

EXAMPLE:Cross-Channel Current Component = 1/2 knotSpeed Through The Water = 8 knotsCourse Correction = 3-1/2o

Source: J. Jussem and DeBoer, "Seven Years Experience With Simulator Training atVLCC Pilots in the Netherlands," Proceedings Second International Conference on

* Marine Simulation. National Maritime Research Center, Kings Point, New York,1981.

sents the single point at which - Number of Rudder Orders prior toownship was closest to the chan- 300 heading changenel boundary. This distance wasmeasured from the skin of the - Number of Course Orders prior to

. ship, not the center of gravity, 300 heading changeto the channel boundary. Theunits of measure were "feet". - Number of Engine Orders prior toIf the vessel went beyond the 300 heading changechannel boundary, a zero valuewas entered. These third-order performance measures

are closely related to the Dutch turn7 Third-Order Performance Measures around technique, which was emphasized

during the experimental training pro-- Course correction at 4 knots -- gram. Four knots was considered as an

the difference between ownships appropriate sample point to investi-course and the channel heading gate the pilot's adherence to the re-when ownship's speed was reduced commended course correction providedto 4 knots. in Table 2 as he slows the vessel.

9,',-, 19

I g•

For the rudder, course, and engine as a result of the training program,orders, an upper limit of 30 degrees (pre versus posttest performance)heading change was chosen because, for yielding multiple measures of trainingthe particular test exercises em- effectiveness. These measures ofployed, this value appeared to define change in performance (i.e., betweenthe limit of the critical region where the two tests for each module) are thethe Dutch technique was most helpful. basis for the findings presented and

discussed in the subsequent ResultsThe data analysis focused on changes and Discussion chapters.in each of these performance measures

"20

9, ,.

'9.

CHAPTER 3

RESULTS

This section presents the results ob- Table 3. The primary performance mea-tained from the CAORF experiment, in sure is the sum of the difference be-terms of the data collected for, and tween the actual ship's position andsignificance of, each of the perfor- the 1 nm radius at each of four datamance measures. The evidence is sum- lines throughout the turn (i.e., seemarized herein, but is not discussed. Figure 4, and the explanation inThe discussion of the results, ad- Chapter 2). This primary performancedressing each of the issues under in- measure, sum of distance off 1 nm ra-vestigation, is contained in Chapter dius, was reduced from a mean of 734.04. The questionnaire data collected to 384.9 (p 0.025), showing a signi-from each of the pilots/subjects are ficant gain in pilot performance.presented and discussed in the Discus- This finding demonstrates the poten-sion chapter. tial effectiveness of advanced instru-

mentation training for pilots on aThe results are segmented under each simulator.of the two training modules -- Ad-vanced Instrumentation, and Emergency The secondary performance measure inShiphandling. Module I addresses the actual rate-of-

turn maintained in comparison with theideal rate-of-turn (0.11 degrees per

3.1 ADVANCED INSTRUMENTATION -- minute). Pilot performance regardingMODULE I this measure was also found to signi-

ficantly increase as a result of theThe results pertaining to this module training program. The mean error inare presented below regarding Training rate-of-turn was reduced from 0.28 toEffectiveness and Pilot Experience. 0.14 (p:O.Ol) as a result of the

training program. Whereas the sum ofTraining Effectiveness distance off the I nm radius repre-

sents the bottom-line of performanceThe training emphasis in Module I fo- in this module, the sum of differencecused on maneuvering a vessel around a between the actual rate-of-turn andconstant radius turn, such that the the desired rate-of-turn representsship's track would maintain an appro- the behavior of pilots in achievingpriate constant radius through the that performance.turn. The constant radius of the turnused during the test was 1 nautical Two additional performance measuresmile (nm). Hence, ideally, a 1 nm describing specific changes in pilot-mile radius would have been achieved ing behavior were also found to change

* during the test scenarios. The sta- significantly as a result of thististical data pertaining to this per- training module (see Table 3). Theformance measure, and five additional mean minimum distance to channelperformance measures, are contained in boundary increased significantly from

21

- ,~- ~ *.-**4 . -' ;..'~ .$ ?' *' 4 , Sx~~'***%.* *.*

>1

4.1

C

CLo C C

0'

IL - -

o1 tooU, 0

4j. 00 to r-.kA. to' 0en(.

0 _ J_ a)

-4~~ 0) --

CC> C C C

C~

00

LU 4.1'oV Q. C-0 0.#a~

-4 C)Cj %o I! C

L&J z,

0. 0) vs W

0- 004 C 0

4w- C n 4 p W 4

*' V V S. c # II

ou zJ 0' 0j c-. -.4 4jt4)ju Io

u i 0 UP J C> c' 0CU 0 4 %6.J ILI

0 ',v R

_j r- 1

4A w In 4V) 0n 0 490 . 1 0>

61~Lf 41 s u1- - In

22C t O

143 feet to 237.5 feet (p 0.025). not result in a significant change inThis finding, of course, would be the bottom-line vessel performance.expected to correlate highly with theprimary performance measure, sum of The limited experience group did notthe distance off 1 nm radius. The show a significant change in perfor-number of course orders issued by the mance on the remaining performancepilots were reduced at a marginally measures. In summary, therefore, asignificant level as a result of the positive, although weak, indication oftraining program, from a mean of 2.25 training effectiveness was observedto 1.17 (p-0.l0). for the limited experienced group of

pilots with regard to advanced instru-The remaining third order performance mentation training.measures did not yield a significantchange in performance as a result ofthis training module. The number and Extensive Experience. The effec-strength of significant results, how- tiveness of advanced instrumentationever, (i.e., four of the six perfor- training for the pilots with extensivemance measures, including the primary experience was found to be strong,and secondary measures) strongly indi- contrary to that observed for thecates that effective pilot training pilots with limited experience, withwas achieved in the advanced instru- several performance measures showingmentation module. significant improvement (Table 5).

The sum of distance off 1.0 nm radius,Pilot Experience the primary performance measure,

showed a significant increase in per-The results pertaining to the impact formance with the mean sum of dis-of experience on the effectiveness of tances of 1.0 nm radius substantiallythe advanced instrumentation training reduced from 821.1 feet to 343.2 feetmodule are summarized in Tables 4 and (psO.05) as a result of training.5, for the limited and extensive ex- The sum of the differences between theperienced groups. actual rate-of-turn and the desired

rate-of-turn, the secondary perfor-Limited Experience. The primary mance measure, was also found to showperformance measure, sum of distance a significant improvement, with theoff 1 nm radius, did not change signi- rate-of-turn error reduced from 0.32ficantly, even though this mean per- degree per second to 0.14 degree performance improved substantially be- second (p O.005).tween the pretest and posttest (626.5feet to 447.2 feet). The secondary Two of the four third order perfor-performance measure, sum of the dif- mance measures were also found to 'showferences between actual rate-of-turn a significant change in performance.and desired rate-of-turn, did yield a The minimum distance to channelsignificant change in performance, boundary increased from 109.5 feet towith a reduction in the mean error 250.0 feet (p.0.025). The number offrom 0.24 to 0.13 degree per second channel excursions was significantly(p!50.05). The behavior of the li- reduced from 4 on the pretest to zeromited experience group in bringing the on the posttest (p:0.06). These re-vessel around the constant radius sults indicate substantial improvementturn, therefore, was found to improve for the pilots with extensive experi-significantly as a result of the ence as a result of the advancedtraining module, even though this did instrumentation training.

23

CL

4141 LA OD C

4

00

Ii - CC 4-:0 41 LA LAtoo4. a~ CDQ L

*4 Mz 4 0

-T CC C

E C C c C C L00

4/)4

Lii L'a4 *l 0O L-

* LA0L 0 LA

x -4

Ci CiJ 0 0to- - 0 9 z 0;

C=

M- Li

W 00

iv = 0 4a 0u Au 0 .C'X v-

V4 0. 0 R

IAA1~~~ ~ mgL C.' q N

-42 Na

00 O~C. N

*jq0

CC

coj

P-4 LO 0CDC.

04i r- '.0 C'.j

4 1) '. C'J

4-4

4.) It

CDC

IA C

V)f

* 1.0(

4AI 0to

)ca(n 0

U.')

U;I

0- CC C CI= 0

CLC

- C

3K a. . aLi

I~ ~ 0n ON qr L

4C 0

4.' S. 4.

~I -

0V) 4

0 0 :.' $- C, #aI

IV . V -Za2 v 0 41C a

*~I 4.) R~~.- i ) V

_ U U oar' Ig.. 4.. .. S4@

4-& 4W J 2, a C .

25

* ~*. ~. .-.-~v;C4

3.2 EMERGENCY SHIPHANDLING -- degree heading change during the turn-MODULE II ing around maneuver was reduced from

7.0 to 4.4 (p0.05); the mean numberThe results pertaining to Emergency of course orders prior to the 30Shiphandling (Module II) are presented degree heading change was increasedbelow with regard to Training Effec- from 1.9 to 4.3 (pf0.05); and thetiveness and Pilot Experience, mean number of engine orders prior to

the 30 degree heading change was re-duced from 6.2 to 3.2 (p!O.02).

Training Effectiveness These three performance measures ad-dress the behavior of the pilots dur-

The data pertaining to the training ing the critical early stages of theeffectiveness achieved during the turning around maneuver (i.e., theemergency shiphandling module are pre- greater part of the turning aroundsented in Tables 6 through 8. The maneuver during which speed is takenprimary, secondary and four third- off, and the ownship heading is sub-order performance measures were cal- stantially changed). The number ofculated to evaluate training effec- rudder orders and engine orders priortiveness. The difference in the fre- to the 30 degree heading change werequency of channel excursions between found to be significantly reduced,the pretest and the posttest was the while the number of course orders wasprimary measure of training effective- found to significantly increase duringness. The frequency of channel excur- this period. The technique addressedsions was significantly reduced from in Module II for turning around empha-six excursions to zero excursions (p sizes incremental changes/modifica-0.02) as a result of this training. tions to ownship course in comparisonThis demonstrates a significant with the channel centerline course,bottom-line gain in pilot performance thus promoting the changes that wereas a result of the training program. observed. These findings further sub-The minimum distance to the channel stantiate the effectiveness of simula-boundary, the secondary performance tor-based emergency shiphandlingmeasure, was found to be marginally training for pilots.significant, but also showing a changein pilot performance representative of Pilot Experiencepositive training effectiveness. Theminimum distance to the channel The training effectiveness data forboundary increased from 75.0 feet to the groups of pilots with limited and129.2 feet (p-50.l0). It should be extensive experience are summarized innoted, the minimum distance to the Tables 7 and 8.channel boundary as used in this in-vestigation is a very conservative Limited Experience. The onlyestimate of positive changes in per- performance measure to reflect a sig-formance due to the manner in which it nificant change for pilots withwas calculated (see paragraph 2.10). limited experience was the minimum

distance to the channel boundary,which increased from 60.0 feet to

Three of the four third-order perfor- 170.0 feet (p-O.05). This significantmance measures showed a significant finding suggests that the module wasimprovement as a result of the emer- effective for training this particulargency shiphandling training. The mean group of pilots. The number of chan-number of rudder orders prior to a 30 nel excursions, the primary perfor-

26

J4"

*L c;;

EU I 0LO LA C\O

.4: co r. L

r- 0n

4-J )

-j C >

a >t O. C C

C C

S- co - o+S-o (a ~o6 r_' ci

0-.. -) i

ce4.' WJ

0 - C Cj CCi Cj

* 0~ C

.- VCoU r

CD.6- . .- C> VEU ~(. U to >-

0*4-)

$. 0- 0-

0 Ju - 0) AA c- 1 4

* v u "D.6

41 c =0 c4-0.

v n C' 4. 41-- *-

USA 0-10 co 0D 04.C) Im 4 *)4.

c L.)f CO 0 C0C I

k 4.' 4

C C

V) 0J Ln

Z; .2;

E C4J

o~ 04)r.4J1

- ~ ~ ~ ~ C CD u~CJ- C

ZiI 4J '0 IV II

CD- LC4-1 C CC CA

CACvI C; C

-A 4J LCDC

Pz. 04-l1=0 CiP) C)L-J

9. ~~~~3 0 a00 'J

LUf toa 41- - C'

4JJ I=. 'acz 40- S 4- ' 0

CL. V)A

fa LOC')OL) I 4JLUj

C 4J04- S- 0_

I-~4 -aa4 * 4 0 C!-~~~4 06' 0 '

4A 0 44 -W

*1 0

c' 1.a 4 - - .- )C10

4. en m. 1.n 4J(D 0 4 -C 4-C %-

1%t en 0A c3o 0 00aC' @'

do 04 M f - 0.c

0 4-. 5C -C =' L.O . C @3Du. 'a m' Z g.. ' 2C C V M 0

28

IN4

ICD

0_ C0a

4-'1

0 CCC..) oe

C -

C IJ

1A . C1 z C" C

C2 R

r, 4A0' m -S m w .c\J

C~ . . c 0 *

00 = C -CVWW u Lu C u C

4. C..)4-b 29

A; G

mance measure, did not show a signifi- pilots with extensive experiencecant difference between the pretest (Table 8). It should be noted, how-and posttests, even though the fre- ever, that the number of channel ex-quency of excursions was reduced from cursions was reduced from three on thethree to zero as a result of the emer- pretest to none on the posttest. Thisgency shiphandling training. This is similar to that found for thelack of statistical significance is pilots with limited experience. Thelikely due to the small number of data fact that this measure was significantpoints in this sample (i.e., five), overall, for both groups of pilotsThis interpretation appears particu- combined as discussed above, stronglylarly appropriate in view of the sig- suggests that the lack of significancenificant results on the conservative for the nearly identical subgroups ismeasure of minimum distance to the due to sample size only.channel boundary. This minimum dis-tance was increased nearly three times Two of the four third-order perfor-between the pretest and posttest, de- mance measures were found to yield amonstrating a substantial change in significant change in performance.performance. None of the third-order The number of rudder orders prior toperformance measures were found to be the 30 degree heading change was re-significant for this group, even duced from 8.6 to 4.7 (p0.05). Thethough their means all changed in the number of engine orders during thisexpected directions as a result of the period was reduced from 6.0 to 2.7training program. (p0O. 10). The significance of these

results suggests that pilot behaviorExtensive Experience. Neither the did change as a result of the trainingprimary or secondary performance mea- program, although this was not demon-sures were found to show a significant strated in the bottom-line primary andchange as a result of Module II for secondary performance measures.

,0

30

1'

CHAPTER 4 V.

DISCUSSION

The results obtained during this in- Hence, it can be concluded that pilots

vestigation yielded a number of signi- may potentially benefit from simula-ficant changes in pilot performance tor-based training.and behavior as a result of the train-ing program. These data may be inter- This finding is extremely important.preted in a number of ways. This sec- It should be cautioned, however, thattion interprets the results in terms this finding does not imply that allof the major issues under investiga- pilots can benefit from any form oftion, and reaches summary conclu- simulator-based training. Rather, itsions. The discussion is divided into shows the potential benefits that canfour major subsections, the first accrue from appropriate training pro-three of which address the simulator grams. The areas in which simulator-data, and the fourth which addresses based training may be useful foran analysis of the pilot opinion ques- pilots is discussed further below.tionnaires. The subsections are: (1)effectiveness of simulator-based 4.2 AREAS OF PILOT TRAININGtraining for pilots, (2) pilot skillareas for which simulator-based train- The analysis preceding the simulatoring would be most effective, (3) the training experiment identified severalexperience level of pilots most likely areas of pilot skill for which simula-to benefit from simulator-based train- tor-based training might be effectiveing, and (4) pilot opinions regarding (see Chapter 2, Methodology). Thethe potential of simulator-based Advanced Instrumentation and Emergencytraining. Shiphandling areas were selected from

the candidate set on the basis of4.1 TRAINING EFFECTIVENESS their potential training effectiveness

within the constraints imposed on thisThe effectiveness of simulator-based experiment. The overall gain in per-training for pilots is the issue of formance as a result of both Module Iprimary concern to this investiga- and Module II training programs, astion. The substantial number of per- discussed above regarding trainingformance measures that show a signifi- effectiveness, demonstrates the poten-cant increase in pilot performance as tial of simulator-based training fora result of training for both modules both Advanced Instrumentation andstrongly demonstrates the potential Emergency Shiphandling. Four of theeffectiveness of simulator-based six performance measures for the Ad-training for pilots. Nine out of 12 vanced Instrumentation module and fiveperformance measures across both of the six performance measures formodules showed significant gains in the Emergency Shiphandling module wereperformance; the primary performance found to show significant performancemeasures in both modules yielded a gains. Furthermore, not only were thesignificant positive training gain. gains significant, but the magnitude

31

-'

of differences were substantial for an area of skill in which pilots couldmost measures. Hence, it can be con- potentially gain from simulator-basedcluded that both Advanced Instrumenta- training. The primary performancetion and Emergency Shiphandling areas measure associated with maintaining acan be effectively addressed for 1.0 nm radius is the summation ofpilots via simulator-based training position errors from that ideal trackprograms. through the turn; as discussed in the

Methodology section, it is calculatedThe finding that effective training on the basis of four data lines spacedfor pilots was achieved on each of the throughout the turn. The ownshiptwo training modules selected suggests position throughout the constant ra-that the highly structured investiga- dius turn was much more closelytive approach used for selection has a aligned with the ideal 1.0 nm radiushigh degree of validity. That is, the of that turn as a result of the train-two areas chosen for this experiment ing program (Figure 5). The mean dis-were selected from a larger group crepancy, in terms of the summed de-similarly identified, and thus to some viations at the four data lines, wasextent are representative of the nearly halved as a result of therationale that was used to select each training program (734 feet versus 384of the potential areas. The strong feet). It provides strong evidencepositive findings in this experiment for the effectiveness of simulator-tend to verify the structured analysis based training in addressing advancedprocess used, and thus further suggest instrumentation aids for piloting.that the other training areas similar- This increase in trackkeeping perfor-ly identified may also have good po- mance provided improved vessel naviga-tential for training benefit from a tion in terms of fixing ownship'ssimulator-based training approach. position. It was achieved as a result

of maintaining the vessel rate-of-turnA further discussion of the findings more closely to the ideal (0.11 de-pertaining to each module of the grees per minute) for this turn; atraining program is presented below, corresponding improvement in the mean

sum of differences between the actualAdvanced Instrumentation rate-of-turn and the desired was

achieved (Figure 6).The benefits accrued as a result ofthe Advanced Instrumentation training A direct benefit of maintaining a con-module are shown in Figures 5 through stant radius through the turn is that9. The performance objective of this the distance to the channel boundarymodule was to maintain a 1.0 nm radius is maximized, placing ownship in anthrough the turn, using available in- optimum position throughout the turnstrumentation (e.g., rate-of-turn in- and minimizing the chance of an excur-dicator, doppler speed log, racon). sion beyond the channel boundary. TheBeing able to maintain a perfect 1.0 minimum distance to the channelnm radius using these instruments boundary, which is a single point per-would provide the pilot with addition- formance measure, should yield perfor-al relevant information under condi- mance similar to the sum of distancestions of reduced visibility (i.e., off the 1.0 nm radius. That is, aownship position through the turn), larger minimum distance to the channelSince these instruments are not com- boundary should be obtained when themonly available on most ships or in constant radius turn (i.e., in thismost pilotage waters, they represented case 1.0 nm) is best maintained. As

32

41

fL. s

1000-PRETEST

800- 387 POSTTEST

SUM OFDISTANCES

OFF 1.0 NMRADIUS

(FEET) 400 3

200

-/1 PC0.025

PRETEST POSTTEST

Figure 5. Module I Training Effectiveness --

Sum of Distance Off 1 nm Radius

.. 0.4-

0.4 Iele~eIPRETEST

kPOSTTESTMEAN 0.3- 0.28

SUM OF0 DIFFERENCES

. 4BETWEEN

" 'V ACTUAL ANDDESIRED 0.2-

RATE OF TURN0:14

0.1-

1o30.001'//

PRETEST POSTTEST

Figure 6. Module I Training Effectiveness --Mean Sum of Differences Between Actual and Desired Rate-of-Turn

33

.1 V4

300 -~7PRETEST

237 POSTTEST t

200 -MINIMUM

DISTANCE TOCHANNEL 143BOUNDARY(FEET)

100-

ji. p5 0O.025

PRETEST POSTTEST

Figure 7. Module I Training Effectiveness --Minimum Distance to Channel Boundary

3.0 -PRETEST

- POSTTEST2.25

2.0-NUMBER OFCOURSEORDERS

1.17

1.0-

PRETEST POSTTEST


Number of Course Orders

34

14-12.9 PRETEST

12 -11.6NSPOSTTEST

10 -0NUMBER OF

RUDDER 8ORDERS 01X

4

2

PRETEST POSTTEST

Figure 9. Module I Training Effectiveness -- Number of Rudder Orders

can be seen in Figure 7, the minimum program was completed, the mean numberdistance to the channel boundary inof course orders used was reduced tocreased substantially as a result of 1.17. This halving of the number oftraining, similar to that which oc- course orders as a result of thecurred for the sum of distances off training program was marginally signi-the 1.0 nm radius. This finding ficant (p!!!0.10). This level of sig-further substantiates the potential of nificance has been accepted in thesimulator-based training of pilots past for this type of research, due towith regard to advanced instrumenta- the relatively low power of the sta-tion. tistical tests resulting from the

highly variable performance dataThe technique used to achieve a con- usually obtained in this type ofstant rate-of-turn maneuver is sensi- experiment. Variable in simulatortive to the initial ownship position training research data can result fromwhen entering the turn, ownship speed, several factors. However, it shouldand maintenance of an optimum rate-of- be noted that a primary factor isturn configured for that particular usually the relatively small number ofturn. Course headings, typically, are subjects used in order to minimizenot relevant in conducting this type cost of the experiment. Based on the

*-of maneuver since the rate is sensi- significance of the primary andtive to amount of rudder. Hence, secondary measures, therefore, thecourse orders would be expected to difference in course orders as a re-decrease, and rudder orders would be sult of the training program is likelyexpected to correspondingly increase, to be significant.The results, although not conclusivein this regard, strongly suggest that Conversely, the number of rudderthese changes in behavior did occur. orders was not found to differ signi-The mean number of course orders used ficantly as a result of the trainingto achieve the turn prior to the ad- program, although the mean value didvanced instrumentation training was increase slightly (Figure 9). Never-2.25 (Figure 8). After the training theless, the proportionate number of

35

rudder orders to course orders appears are presented in Figures 10 through 13to have changed as expected, suggest- and discussed below.ing that the pilots modified theirbehavior in transiting the constant The six channel excursions resultingradius turn as a result of the simula- from the pretests shows the difficultytor-based training program. of turning a large vessel around in a

relatively narrow channel. TheIn summary, the Advanced Instrumenta- absence of any channel excursions ontion training program resulted in the posttest demonstrates the effec-changes in pilot behavior and perfor- tiveness of simulator-based trainingmance as expected, achieving improve- in polishing the skills of experiencedment in shiphandling performance per- pilots. The turning around maneuvertaining tc constant radius turn tech- was accomplished by using a specificniques. he empirical findings show technique, which has been developed by

* that significant benefits for pilots Dutch pilots in the port of Rotter-may exist as a result of participation dam. The effectiveness of this rela-in simulator-based Advanced Instrumen- tively brief (only 8-hours long)tation training, training program atests to the poten-

tial effectiveness of simulator-basedEmergency Shiphandling training for pilots, as well as the

high-level of skills and knowledgeThe simulator-based training program possessed by the pilots participatingwas found to be effective for the type in this experimental program. Theseof emergency shiphandling addressed pilots were able to rapidly learn and

during this program, as evidenced by apply the Dutch technique in a diffi-the significant changes in performance cult shiphandling situation. Commer-summarized in Table 6. The bottom- cial training programs would certainlyline measure of shiphandling perfor- spend considerably more time address-mance is the number of collisions, ing and practicing appropriate emer-rammings, and groundings, and in terms gency shiphandling, as do the Dutchof this experiment, the frequency of pilots during their week long re-channel excursions. The turning fresher training program at Wageningen.around maneuver that was addressed inthis module was believed to be quite The minimum distance to the channeldifficult, to the extent that a signi- boundary during the turn around maneu-ficant potential for going beyond the ver was found to improve in directchannel boundaries existed. Hence, correspondence to the number of chan-this became the primary performance nel excursions, increasing from 75measure. The significant difference feet to 129 feet (Figure 10). Thisin the number of channel excursions represents an increase of 72% as abefore and after the training (i.e., result of the training program. Al-six excursions before, none after- though the level of statistical signi-wards) alone attests to the potential ficance of this increase is marginaleffectiveness of simulator-based (p!O.l0), the method of calculationtraining for this area of pilot together with the number of channel

skills. The number of other signifi- excursions suggest that the obtainedcant performance measures also showing level of statistical significance is aa training gain (i.e., 3 of the re- conservative estimate. The computedmaining 5) further substantiates this mean distances were based on the actu-finding. A summary of the findings al distance for each vessel that didpertaining to emergency shiphandling not go beyond the channel boundaries,

36

*41~..~ ~* *. - - ., - ~ . ~' 4~WWI;,*

150-

129

SPRETEST100-

MINIMUM DISTANCE NNNIPOSTTESTTO CHANNEL

BOUNDARY (FEET) 75

501

PRETEST POSTTEST

Figure 10. Module 11 Training Effectiveness -

Minimum Distance to Channel Boundary

8Z PRETEST7.0

79 POSTTEST

NUMBER OFRUDDER ORDERS

PRIOR TO 3004.HEADING CHANGE 4

2 p!!: 0.05

PRETEST POSTTEST

Figure 11. Module II Training Effectiveness -

Number of Rudder Orders

37

5-r

16 1 PRETEST

4.3

4- POSTTEST

NUMBER OF

COURSE ORDERS 3

PRIOR TO 30"

HEADING CHANGE 1.9,

2 :

1 p < 0.05

PRETEST POSTTEST

Figure 12. Module II Training Effectiveness --

Number of Course Orders

;7 - PRETEST

6.2

NUMBER OF 6- POSTTEST

ENGINE ORDERS

PRIOR TO 300 -

-HEADINGCHANGE 4 -

3.2

2 - \2, p < 0.02

PRETEST POSTTEST


Number of Engine Orders

3-

, ~38 ;

*', , "v '. -, -., " ",- -, -.'. " '.-. '.\ ...., " * ' '.'. . ',' ',.' \. L ; , , , ... " .

and a value of "szero" for those yes- on the posttest (Figure 11). Corre-sels that went beyond the channel spondingly, the number of courseboundary. Thus, these mathematical orders were significantly increasedcomputations included the ratios of from 2.2 on the pretest to 4.1 on thedistances for vessels staying within posttest (Figure 12). These findingsthe channel, while adding only a zero are consistent with the Dutch tech-(i.e., rather than a negative value) nique employed, and as expected. Theyfor those that went outside the chan- provide the rationale for improvednel. The resulting performance mea- pilot performance in turning the largesure, therefore, is weighted in favor vessel around in a narrow channel, inof providing a within-channel value, terms of changes in pilot shiphandlingHence, the difference between pretest behavior during that maneuver.and posttest performance as calculatedfor this measure would be conserva- The early stages of the Dutch turningtive. This marginally significant around technique focus on balancingresult should thus be interpreted as the ship's cross channel velocity vec-definitely significant. The simulator tor with that of the current, andwas found to be an effective medium hence require the pilot to modify thefor training emergency shiphandling ship's heading as a result of changestechniques to pilots. in ownships speed. As the speed de-

creases, greater heading changes areThe turning around technique developed made to counter balance the current,by the Dutch places considerable em- and thus to ideally keep the center ofphasis on the early stages of the ownship in the center of the channelturning around process. During the (note, course would ideally remainearly period the speed of the vessel along the channel centerline). Ifshould be reduced and the vessel's sufficient distance ahead is avail-heading changed appropriately in cor- able, little if any use of the enginerespondence with the procedure out- during these early stages is neces-lined in Table 2. A 30 degree heading sary. Rather, the engine is stopped,change with respect to the channel and as the ship's speed falls off overcenterline was chosen in this experi- time course corrections are made toment to represent the point at which counter balance the current. It wasthe early stages of the turn around expected, therefore, that the trainingmaneuver were concluded. This was program would result in fewer engineused as a reference datum for evaluat- orders during the early stages of the

*ing performance. The turning around turn around maneuver. A significanttechnique relies on a difference be- reduction in the number of enginetween the shipheading and the channel orders given was, in fact, obtainedcenterline's course in correspondence (Figure 13). The mean number ofwith the current's magnitude and di- engine orders dropped from 2.25 priorrection, and ownship speed. Hence, it to the training program to 1.17 fol-was expected that the number of rudder lowing the training program. Thisorders given would decrease, and the findi-ng further describes the changesnumber of course orders would con- in pilot behavior that led to improvedversely increase, as a result of the turning around performance.training program. The number of rud-der orders issued prior to a 30 degree In surmmary, the Emergency Shiphandlingheading change was significantly re- training module was found to be effec-duced as a result of the training pro- tive. Significantly fewer channelgram, from 7.0 on the pretest to 4.4 excursions occurred after the training

~~39

V5'A:

program, with the pilots using corre- cations should be viewed as approxi-spondingly less of the channel width mate.to turn the vessel around. These

Q emergency shiphandling performance The analysis investigated each of theimprovements were directly related to two pilot experience groups indepen-observed changes in pilot behavior as dently, with regard to each of the twoa result of the training program. training modules. Particular caution

* should be exercised in interpreting4.3 LEVEL OF EXPERIENCE the findings relating to experience,

since relatively small sample sizesThe third major issued addressed by were used in each group. The resultsthis experiment is the difference in are discussed below regarding Advancedthe effectiveness of simulator-based Instrumentation and Emergency Ship-

N training for different groups of handling.pilots. The experiment investigatedpilots with limited experience (i.e., Advanced Instrumentation I

* less than 5-years) and pilots withextensive experience (i.e., greater The pilots with extensive experiencethan 10-years). These two categories appear to have benefited more from theof pilot experience, although appear- Advance Instrumentation training thaning straightforward, are actually did those with limited experience.somewhat complex. The differences in The differences in training effective-entry qualifications and training pro- ness for the two groups are reflectedgrams for the various pilot organiza- in the three performance measures pre-tions make it difficult to categorize sented in Figures 14 through 19. Thethe amount of experience a pilot has, track radius error showed a signifi-For example, some pilots enter the cant positive training gain for thepiloting profession after having had a extensive experience group, while aconsiderable amount of experience as a similar result was not obtained fordeck officer (i.e., on a sea-going the limited experience group (Figurevessel, a tug boat, etc.). Although 14). This finding demonstrates thanthese individuals may have had only the Advanced Instrumentation traininglimited experience as a pilot since program was effective for the experi-joining the pilot organization, they enced group of pilots. The lack of amay have had considerable experience significant finding on this measureas deck officers and in working with for the limited experience group does

* pilots. Other pilot organizations, at not show that the training was inef-the opposite extreme, do not have any fective; rather, it indicates that theprerequisite experience requirements data were insufficient to demonstratefor entering the piloting profession. an effective gain. It should be notedThese organizations, however, typical- that the limited experience group didly have a very extensive training pro- achieve a relatively large differencegram spanning a number of years. The in performance (i.e., 626.5 feet on

* pilot/subjects within both the limited the pretest versus 447.2 feet on theand extensive experienced groups re- posttest), even though it was notpresented both types of backgrounds, found to be statistically signifi-although they all met the established cant. This lack of statistical signi-criteria for years experience as a f icance may be due to the power of thelicensed pilot within their assigned test, resulting from the relativelygroups. As a result, these classifi- small sample size. A relatively small

40

eq°

PRETEST

LXIPOSTTEST* 1000

821.1

" 800 :

SUM OF

DISTANCES 600 -OFF 1.0 NM

RADIUS (FEET)400 343.2

NON 0.05

200 SIGNIFICANT

LIMITED EXTENSIVEEXPERIENCE EXPERIENCE

Figure 14. Level of Experience, Module I--Sum of Differences off 1.0 nm Radius

~ PRETEST0.32

0.3- POSTTEST

0.24

SUM OF - •

DIFFERENCES 0.2BETWEEN 0- -

ACTUALAND DESIRED 0.140 .13 "RATE OF TURN 0.13

0.1- P 0.05 p!0.005

,..1LIMITED EXTENSIVE

EXPERIENCE EXPERIENCE

Figure 15. Level of Experience, Module I --

Sum of Differences Between Actual and Desired Rate-of-Turn

41

%

300

SPRETEST250.0

MINIMUM DISTANCE tZ JPSTSTO 200-

* CHANNEL BOUNDARY(FEET)

NON 109.5100 SIGNIFICANT

p -c 0.025


Figure 16. Level of Experience, Module I -


300- PRETEST

SPOSTTEST

MINIMUM DISTANCETO 200-

CCHANNEL BOUNDARY 170* -~ (FEET)

100-NNp05

SIGNIFICANT ~ 50060


Figure 11. Level of Experience, Module II -


42

4..?

14 PRETEST

12-

POSTTEST

10 8.6

NUMBER OF 8RUDDER ORDERS

6- 4.7NON

SIGNIFICANT p <0.054-

2-

LIMITED EXTENSIVE

EXPERIENCE EXPERIENCE

Figure 18. Level of Experience, Module II --Number of Rudder Orders

10- PRETEST

8- ~POSTTEST

NUMBER OF 6 6.0

ENGINE ORDERS

4-NON 2.4

SIGNIFICANT p0.1

2


Figure 19. Level of Experience, Module II --Number of Engine Orders

43

1~ %-

number of pilot/subjects were used in groups appear somewhat opposite tothis condition due to cost and experi- those found for the advanced instru-mental design considerations. This mentation module. The pilots withresults in relatively low statistical limited experience appear to havepower for these types of real-world benefited greater from this moduleexperiments, and greatly increases the than did those with extensive experi-chances of not obtaining statistically ence. The primary performance mea-significant results even though they sure, frequency of channel excursions,do exist. did not show a significant difference

for either group separately as a re-Both groups of pilots achieved a sig- sult of the training program. Sincenificant positive training gain in the this measure was found to be statisti-rate-of-turn error, a secondary per- cally significant with both groupsformance measure (Figure 15). These combined, it can therefore be con-results support the above contention cluded that the lack of significancethat the training program was effec- for either group when separated is duetive for both groups, and that the to the small number of pilots in eachlack of significance of the substan- group and the resultant low statisti-tial track radius error reduction was cal power.due to the power of the statisticalanalysis. The limited experience pilots signifi-

cantly improved their minimum distanceThe minimum distance to the channel to the channel boundary as a result ofboundary was found to be significant the training program (Figure 17).for the extensive experience group, This measure is closely related to thebut not for the limited experience number of channel excursions. Thisgroup (Figure 16). The mean minimum group of pilots on the posttest nearly

. distance to the channel boundary did tripled their pretest minimum distancenot change appreciably for the limited to the channel boundary (i.e., 60 feetexperienced group, hence the lack of to 170 feet). The pilots with exten-statistical significance. This find- sive experience, on the other hand,ing supports the above contention that achieved a relatively small increasethe extensive experience group bene- in the minimum distance to the channelfited more from the Advanced Instru- boundary (86 feet to 100 feet), whichmentation module than did the limited was not statistically significant. Itexperience group. is concluded on the basis of these

findings that the pilots with limitedIn summary, the experimental results experience benefited more from theindicate that both groups of pilots Emergency Shiphandling module than didbenefited from the Advanced Instrumen- the pilots with extensive experience.tation training module. The extensiveexperience group, however, was found The extensive experience pilots did,to benefit more from this training interestingly, change their behaviorthen the limited experience group. as a result of this Emergency Ship-The reasons for this difference are handling module, although it was notnot apparent from the data collected. reflected in a statistically signifi-

cant change in vessel performance.Emergency Shiphandling The number of rudder orders and the

number of engine orders used by thisThe training benefits of the Emergency group was significantly reduced as aShiphandling module for the two pilot result of the emergency shiphandling

44

e? 4"I

training (Figures 18 and 19). These tween the two groups strongly sug-findings are in direct correspondence gests, as was expected, that the bene-with those for the total group of fits to be derived from the differentpilots, as discussed previously; and areas of simulator-based training ofthey are in the direction expected. pilots will depend on their back-These findings suggest that the tech- grounds, such as level of experience.nique used to achieve the vessel turn This finding was not unexpected, sincearound by the extensive experience the benefits to be derived from anygroup of pilots changed as a result of training program depend heavily on thethe training program. Their level of backgrounds of the particular stu-activity reduced considerably in terms dents. Hence, it is likely that

*of both rudder orders and engine apprentice pilots, limited experiencedorders. Furthermore, it should be pilots, and extensively experiencednoted that although a significant dif- pilots will derive differential benie-ference in the distance to the channel fits from the simulator-based trainingboundary was not achieved, the mean in the various areas of pilotingdifference that did occur was in the skill. The extent of benefits to bepositive, rather than negative, train- achieved from such training by theseing gain direction. Hence, as a re- groups, however, cannot be determinedsult of these significant changes in on the basis of these data.

* behavior, it is concluded that thepilots with extensive experience also 4.4 PILOT OPINIONSshowed training gain as a result ofthis module. Pilots in the United States have re-

ceived relatively little exposure toIn summary, the above findings suggest simulator-based training programs.differences in the potential benefits Pilots have participated in simulatorto be accrued from simulator-based experiments, but these are substan-training on the basis of the pilot/ tially different than a structuredstudent level of experience. Although training program designed specificallyboth groups of pilots showed benefit to address the improvement of pilotingfrom each of the two training modules, skills. A broad range of opinionthose with extensive experience ap- exists in the piloting community re-peared to gain more from the Advanced garding the potential effectiveness ofInstrumentation module, while those simulator-based training. An impor-with limited experience appeared to tant part of the empirical data col-

*gain more from the Emergency Ship- lected as a part of this experiment,handling. The findings are not strong therefore, was the opinion of pilotswith regard to the differences between with regard to the potential of simu-the groups, probably a result of the lator-based training as a result ofsmall sample size used. Furthermore, their having participated in a struc-the pilots in both groups, although tured training program. For that rea-

4',technically fitting into the respec- son, the pilots participating in thistive categories, differ substantially experiment were viewed not only asin their backgrounds. Hence, a con- subjects, but also as evaluators.clusion cannot be reached as to which Discussions concerning the variousgroups of pilots will definitely bene- aspects of the training program werefit mare from either the Advanced In- carried out during the free moments ofstrumentation or the Emergency Ship- the course. In addition, a debriefinghandling training. On the other hand, questionnaire was administered to eachthe number of differences found be- pilot individually at the conclusion

45

of the training program. The intent modules; and (4) the general potentialwas to collect as much information as of simulator-based training forpossible regarding the pilots' view pilots. Sixteen of the 34 questionspoints, insights, recommnendations, etc. address the general value and poten-

tial of simulator-based training forA production training facility would pilots. These are the issues of pri-normally modify various aspects of the mary concern to this investigation,program (i.e., including simulator, and are discussed in detail. The re-modules, etc.) after each time the maining questions address specific

* course is presented, on the basis of details of the particular experimentalinputs from instructors and students. training program provided. These de-

* This, of course, is the approach the tails are of substantial importanceRotterdam pilots followed after first from the standpoint of designing andinstituting simulator-based training, conducting pilot training. A briefFor example, the Dutch spent approxi- overview of these is provided. If themately two years in refining the training program material developedhydrodynamic characteristics of the for this investigation were to be usedship models they were using, to make to provide additional training to

* them acceptable to the pilots. Due to pilots, these latter observationsthe experimental nature of the train- would be extremely important to theing program used in this investiga- improvement of that program. All of

* tion, it remained the same for the the information collected from thisthree times it was presented (i.e., questionnaire has been carefully ana-thus maintaining the necessary degree lyzed with regard to the developmentof experimental constancy across the of the training system guidelines forthree groups). The information sum- pilots which is a separate documentmarized herein provides an overview of being developed under this project.the pilots' observations and opinions.Th meodlg framistton fThe questionnaire analysis and discus- the questionnaire was presented pre-sion is presented below in two parts. viously in the methodology section.The first part provides a summnary of It should be recalled that the ques-the questionnaire data, with regard to tionnaires were filled out by an ana-the specific issues being addressed. lyst during a debriefing session withThe second part focuses on the larger each individual pilot at the comple-issues of interest to this investigation of the week's training programtion (i.e., potential of simulator- and experiment. The views expressedbased training for pilots), integrat- by the pilots in the questionnaireing the questions and responses as reflect not only their experiences asappropriate. students/evaluators in the training

program, but also as evaluators whoQuestionnaire Suninary spent considerable tm disues in

simulator-based trainingisu.* The questionnaire addressed a variety

of issues pertinent to the investiga- A summnary of the specific responses oftion, including (1) benefits gained the pilots to the 16 primary questionsfrom each of the three training addressing the value and potential ofmodules; (2) the potential for pilot simulator-based training is presentedtraining regarding each of the three in Table 9. The questions and re-modules; (3) problems and recommended sponses are grouped according to eachimprovements for each of the three of the three training modules and a

46 .-

*~ a %: 44

* TABLE 9. PILOT QUESTIONNAIRE SUMMARY DATA FOR SELECTION QUESTIONS*

*Question** Responses

Advanced Instrumentation

1. Module rating (1 --1 highest) 7.58/10 (mean value)5. Learned something valuable? 11/12 (yes)7. Valuable for apprentices? 11/12 (yes)8. Valuable for experienced pilots, periodically? 11/12 (yes)

Emergency Shiphandling

11. Module rating (1--&l0 highest) 6.58/10 (mean value)12. Is this valuable? 11/12 (yes)14. Valuable for apprentices? 11/12 (yes)15. Valuable for experienced pilots, periodically? 9/12 (yes)

Decisionmaking

19. Module rating (1--I highest) 6.29/10 (mean value) (8pilots)

20. Learned something valuable? 3/8 (yes)22. Valuable for apprentices? 7/8 (yes)23. Valuable for limited experience pilots? 8/8 (yes)24. Valuable for experienced pilots, periodically? 3/8 (yes)

General ____________________ ____________

27. Simulator potential for apprentices? 11/12 (yes)28. Simulator potential for licensed pilots? 11/12 (yes)32. Beneficial training in generic port? 11/12 (yes)

simulator-based training.

* **Question number and brief description, from debriefing questionnaire

(Appendix A).I

47

general category. All of the pilots further enhance pilot skills. Theparticipated in the Advanced Instru- use of simulator-based trainingmentation and Emergency Shiphandling should be voluntary, rather thantraining modules. Eight of the 12 mandatory.pilots were able to partially partici-pate in the decisionmaking training The final two points in this list weremodule, while one group of four pilots expressed by most pilots. Theywere unable to participate in this strongly approve of their current

Amodule due to simulator problems. training programs, but do feel thatHence, the questionnaire data pertain- value may be had f rom simulator-baseding to this module is not considered training as discussed below. Theyto be of the same quality as that of further feel that the positive bene-the first two modules. fits to be potentially gained from

such training would be negated if itA substantial number of specific were forced upon them. The pilotsissues were raised by the pilots believe that the need for simulator-regarding many of the questions. based training and its potential valueThese issues dealt with qualifications is a local port issue, depending on a

*to their answers, generally regarding variety of factors (e.g., current(1) conditions under which their ans- training program, port characteris-

*wers would be given one way or another tics, issues to be addressed by simu-(i.e., yes versus no); and (2) policy lator-based training, individual pilotissues pertaining to how such training backgrounds, and so on). When simula-would come about. Several examples of tor-based training is provided it

*caveats expressed by pilots are: should be tailored to the needs of aparticular port. The final issue is

*The piloting problems and instru- the financing for such training. Thements addressed should be specific use of simulator-based training byto each pilot's port area. pilots will directly depend on the

cost of such training, and the avail-9 The instrumentation addressed able financing.

should a represent real breakthrough, rather than commonly Training Program Value and Potentialavailable instruments.

The questions addressing the value and.1 The emergency shiphandling tech- potential of simulator-based training

niques addressed should represent were generally concerned with: (1)new techniques, rather than tech- the value of that module as experi-niques with which the pilots have enced by the pilots; (2) the pilots'experience, perceived value of that module for

apprentice training; and (3) theo Although the content of simulator- pilots' perceived value of that module

Abased training has good potential, for periodic refresher training ofits actual value would depend on licensed pilots. The responses tolocal availability and cost. these issues are presented in Table 9.

e Simulator-based training would not Advanced Instrumentation. The Ad-be an acceptable substitute for vanced Instrumentation module wasaspects of current pilot training evaluated by the pilots as havingprograms; rather, its value is as a value as conducted, and as havingsupplement to current programs to potential for both apprentices and

48

experienced pilots. The pilot ratings ticular instruments, the local condi-of this course had a mean value of tions in their particular ports (e.g.,7.58 out of a possible 10 points, availability of racons), and so on.This above average rating reflects the These findings demonstrate the poten-value of the Advanced Instrumentation tial value of Advanced Instrumentationmodule as presented. Each of the simulator-based training; they do nottraining modules, as noted previously, address the specific value of thiscan and should be improved as a result type of training for any particularof each application. The ratings port or group of pilots. The specificgiven to this and the other two value would, of course, depend on themodules reflect the initial presenta- particular aspects of Advanced Instru-tion of each; improvements could cer- mentation addressed, as well as thetainly be made as a result of the pilot and port characteristics.pilots recommendations and the staffrecommendations stemming from this Emergency Shiphandling. The pilotinitial conduct of the training pro- opinions regarding the value and po-gram. The high proportion (i.e., ll/ tential of the Emergency Shiphandling12) indicating something of value was module were similar to those of thelearned from this module further sub- Advanced Instrumentation module. Thestantiates the rating. Of greater mean rating given to the module asimportance to this investigation, how- taught was 6.58 out of a possible 10ever, is the potential for such train- points (see Table 9). This aboveing. A very high proportion of the average value is supported by the highpilots stated that the Advanced In- proportion that indicated they learnedstrumentation type of training had something valuable from the emergencysome value for both apprentices and shiphandling module (i.e., 11 out ofexperienced pilots (i.e., 11/12 and 12 pilots). A high proportion of the12/12 respectively). These opinions pilots felt that this type of traininare in good correspondence with the would be of value for apprentices (1performance data collected regarding out of 12), while two-thirds of (9 outthe training effectiveness of the Ad- of 12) indicated that periodic Emer-vanced Instrumentation module. gency Shiphandling training would be

of value for licensed pilots. This isThe piloting skills and knowledge ad- the type of training that the Rotter-dressed under the Advanced Instrumen- dam/Europort pilots participate intation module are not in common prac- every two years.tice in most ports, and hence werebelieved to have potential training On the basis of the questionnaire re-value. The findings, both performance sponses it can be concluded thatand opinion data, substantiate that pilots believe simulator-based Emer-Advanced Instrumentation training does gency Shiphandling training has valuehave value and potential. It should for both apprentices and experiencedbe cautioned, however as noted above, pilots. This conclusion correspondsthat these findings pertain to the well with the performance data fromgeneral potential of such training, the training program, which demon-As several of the pilots stated, the strated an improvement in Emergencyrelevancy of Advanced Instrumentation Shiphandling skills as a result of thetraining, and hence the benefits to be training. Of the three pilots thathad by any particular group of pilots, felt the Emergency Shiphandling train-will depend on the degree of experi- ing would not be of value to experi-ence the pilots have had with the par- enced pilots, one said it was because

p. 49

the techniques were not new to him; the other hand, the pilots generallyanother stated it would depend on the felt that the Decisionmaking modulesituation addressed. If the appropri- would not be of benefit as refresherate issues are addressed during this training for experienced pilots (i.e.,module, therefore, the proportion of three out of eight pilots felt itpilots supporting Emergency Ship- would be of benefit). Hence, it canhandling for licensed pilots would be be concluded on the basis of thesimilar to that supporting the Ad- pilots opinions that the Decisionmak-vanced Instrumentation training. ing module has potential value for the

training of apprentices and pilotsDecisionmaking. The pilot opinions with limited experience. It is notregarding the Decisionmaking module believed to have very much benefit forwere found to differ from those of the the training of experienced pilots.previous two modules. A difference of These findings cannot be further sub-opinion was expected, since it was stantiated with empirically derivedbelieved that the Decisionmaking data, since the data were not avail-

*module would be most appropriate for able from the experiment. Further-pilots with limited experience, and more, the findings regarding thisperhaps also for apprentice pilots, module should be treated cautiouslyFurthermore, only one group of pilots due to the relatively small samplecompleted the Decisionmaking module; a size (i.e., eight pilots) and the dif-

*second group of pilots completed the ficulties encountered in conductingmajority of the Decisionmaking module, this training.while the third group of pilots didnot participate at all in this General Potential. Responses to themodule. Simulator problems caused three major questions pertaining tothis discrepant application of the the general potential of simulator-third module. It is likely, there- based training for pilots (i.e., ques-fore, that the pilot opinions ex- tions 27, 28 and 32) are included inpressed regarding this module are af- Table 9. These questions addressedfected by the manner in which it was the potential of such training forpresented during the program. apprentices and licensed pilots, and

also addressed the potential of con-The rating given to the Decisionmaking ducting such training in a genericmodule by the pilots had a mean value port data base. Whereas the precedingof 6.29 out of a possible 10 points, questions addressed the specificEight pilots answered the questions modules presented during the trainingregarding this module. Although this program and the potential for similar

4rating is above five on a scale of one modules in a production training pro-to ten, it is also the lowest of the gram, these latter questions addressedthree modules. Only three of the the broad issue of simulator-basedeight pilots felt that they had training without regard to any parti-learned something valuable from this cular modules. A large proportion ofmodule. On the other hand, a high the pilots (i.e., 11 out of 12) feltproportion of the pilots felt that that simulator-based training had po-this type of training would be of tential for both apprentices and li-value for apprentice pilots (i.e., censed pi lots. As noted above, how-seven out of eight pilots), and vir- ever, this statement should be inter-tually all of the pilots felt it would preted cautiously. The pilots hadbe of value to pilots with limited many specific concerns regarding theexperience (i.e., eight of eight). On effective application of simulator-

50

based training, several examples of of current relevance in their respec-which have been presented earlier in tive port areas. For example, severalthis subsection. Hence, their support pilots expressed disinterest in learn-for such training in general should be ing about the use of racons sinceviewed to indicate that the potential these devices are not currently avail-is good if it is used appropriately, able in their port area. Operational

training programs, using a genericThe geographic data base used during port data base, should certainly focusthis experiment was that of a generic on those issues of local interest toport, Lotsedorf. The characteristics the pilots being trained. The opinionof this port were specifically devel- was generally expressed that theaped to create the desired training generic port concept is good if thatencounter situations; to enable the generic port has characteristicsconduct of a structured training pro- (e.g., bank effects; racons; etc.)gram, and to enable the collection of that are relevant to the particulardesired empirical data. It should be pilot's port. Hence, valuable train-noted that this port was not developed ing could be provided in the genericto be fully representative of the port context, when addressing issuesrange of piloting conditions likely to of local concern. This finding sug-be encountered by the participating gests that a single well-developedpilots in their respective pilot re- generic port data base may be adequategions, and certainly is not represen- for effective training use by mosttative. Nevertheless, the Lotsedorf pilots, substantially reducing theport was believed to embody sufficient potential cost of such training.characteristics as part of thisexperimental training program to be Specific Training Program Contentrepresentative of the generic portconcept for pilot training. The simu- A variety of comments and recommnenda-lator-based training of pilots in tions were made by the pilots regard-their own port data base would cer- ing the specific content of the expe-tainly have value beyond that of a rimental training program conducted atgeneric data base, although the latter CAORF. The detailed response data arewas also believed to have significant contained in Appendix A. Severalvalue for training pilots. The costs specific findings are discussed below.associated with developing a largenumber of specific port data bases to Advanced Instrumentation. The Ad-Itrain various groups of pilots would vanced Instrumentation module, aslikely be prohibitive, thus making the noted earlier, was generally wellgeneric port concept an attractive received. Several pilots indicatedalternative. Furthermore, for the they felt it had the greatest poten-purposes of this experiment pilots tial for simulator-based training.from a variety of ports were brought Several pilots stated that the lengthtogether, this also necessitated a of this module, as conducted, was in-generic port data base. Most of the sufficient. This is certainly likelypilots (11 out of 12) supported the to be the case, since the module wasuse of a generic data base for the developed to address a relativelytraining of both apprentice and ex- large amount of material in a briefperienced pilots. It should be noted, period of time, due to experimentalhowever, that several pilots were cri- considerations. A production trainingtical of the training program's par- module that would address advancedtial focusing on issues that are not instrumentation should probably span a

51

1. .7 9 . '.VI -... V * -

* ~ ' * N'a, - '%'~11k,

considerably longer period of time. technique addressed was either notPilot commnents regarding insufficient their concept of Emergency Ship-length addressed both the classroom handling, unrealistically easy, or tooand the simulator sessions. Addition- clinical. Several pilots felt veryal comments about this module were: strongly about the value of simulator-

based Emergency Shiphandling training,* Several pilots criticized the both in the form provided during this

module as not addressing issues program, as well as with regard tothat are directly pertinent to other aspects of Emergency Ship-their port (e.g., racons are un- handling not addressed by this pro-

*available in their port). gram. Several pilots suggested thatthis is the major area of potential

* Improvements to the course recoin- benefit for the simulator-based train-mended by the pilots included: (1) ing of pilots. Several comments re-bringing the ship all the way into garding this module included:the port at once, as opposed todoing it in segments; (2) providing * The simulator provides a safe situ-more situation-type of experiences, ation for developing skills andas opposed to the more sterile trying new techniques.practice exercises that were ad-dressed; (3) providing more prac- * More time should be spent on thetice time; (4) providing more dif- bridge, rather than in the class-ficult scenarios; and (5) providing room; the converse was also cited.more port-specific problems (i.e.,not geography specific, but condi- *Several pilots felt the programtions more specific to their local length was about right, whileports). others felt it should be expanded

to address other types of emergency9The pilots recommended additional situations.areas of Advanced Instrumentationthat simulator-based training might e Several pilots felt that the turn-address: (1) advanced portable ing around technique took too muchelectronic navigation aids, such as time; they would like to see morethe Portable LORAN Assist Device rapid and varied situations.(PLAD); (2) advanced radar plottingaids (ARPA); (3) bow thruster indi- e The ship's handling characteris-cator; (4) controllable pitch pro- tics, as modeled, were generallypeller; (5) portable fathometer; felt to be too good for developing(6) advanced radar techniques; (7) the emergency shiphandling skills

*radio direction finder; and (8) use addressed.of Advanced Instrumentation withbank effects. e Several course improvements were

recommended: (1) both shorter andEmergency Shiphandling. Most pilots longer program lengths; (2) a widergenerally felt that the Emergency variety of emergency situations, asShiphandling module was worthwhile as well as exposure to a larger numberconducted and that the Dutch turning of situations; (3) shorter durationaround technique was useful. Several situations; and (4) a range of as-pilots, however, expressed the opinion sociated conditions under which tothat the particular turning around

52

practice the maneuvering, such as a More local problems were requestedpoor visibility, a more narrow by several pilots.channel, more current, and timeconstraints.

* The course improvements recommendede Additional areas of Emergency Ship- included: (1) having only partial

handling training that were recoin- astern rpm available; (2) use of amended included: (1) comparison of poor handling vessel for ownship;average and poor handling ships in (3) comparison of various shipsimilar scenarios; (2) use of emer- types and handling characteristics;gency anchoring; (3) addressing and (4) use of more situations thatrudder failure conditions; (4) ad- require stopping.dressing power failure conditions;(5) creatioh of scenarios that re-quire very quick thinking; (6) ad- # Additional areas of Decisionmakingdressing crosscurrents with course that were recommended included:changes; (7) addressing winds; (8) (1) mechanical breakdowns (engine,requiring anchoring in a short steering); (2) techniques for stop-time; (9) addressing speed together ping the ship completely; (3) awith anchoring; (10) addressing situation with a small boat runningbank effects; (11) addressing ship! across ownship's bow; (4) looseship interaction; (12) turning steering; (5) quartermaster error;around in an anchorage; and (13) (6) bank effects; (7) abrupt divesaddressing fog onset with oncoming toward the bank, shoal, and dock;traffic. and (8) having a meeting ship lose

power and/or steering.Decisioninaking. The pilots had re-latively poor experience with the De- The above observations provide a num-cisionmaking module due to the simula- ber of recommendations relevant to thetor problems. Nevertheless, a number further development of the threeof observations were made by eight of training modules addressed in the ex-the pilots that had some experience perimental training program. Commer-with this module. They include the cial training establishments, as wellfollowing: as pilot groups, would benefit from

the careful consideration of these* Several pilots felt this module was comments with regard to the develop-

good, since it addressed specific ment of training programs for pilots.situations in a realistic context.

Simulator-based Training Potential* The ownship handling characteris-

tics were called out as being too The major issues of concern to thisgood, and hence somewhat unrealis- investigation have been: (1) the ef-tic. fectiveness of simulator-based train-

ing for pilots; (2) the pilot skille Several pilots felt that this areas for which simulator-based train-

module should be combined with the ing would be effective; and (3) the- ~ emergency shiphandling module, such pilot experience levels for which

that the emergency training would simulator-based training would be ef-Ibe placed in the context of the fective. The empirically derived re-situations that occurred in this sults agree well with the question-module. naire results on each of these issues.

53

'~ ~ ~ L

* The pilots participating in this result of the cost/effective natureexperiment generally felt that of constructing an experiment. Thesimulator-based training does have primary intention of this programpotential for both experienced and was to generate as much useful in-apprentice pilots, formation as possible from the ex-

periment, necessitating allowance* The content of the training pre- of only the minimum amount of time

sented to pilots should address for the various piloting issuesissues of local interest, although addressed during training. A com-not necessarily in a local geo- mercial training program shouldgraphical -context; that is, the spend considerably more time ad-particular issues, conditions, dressing the issues of each module.etc., should be of specific localinterest, although the port could * The pilots also indicated that thebe generic, use of simulator-based training

should be to augment their currente Different types of training should training programs, and not simply

be provided to pilots of differing be used in place of certain aspectsexperience levels (e.g., appren- of their current training. Thistices, limited experienced pilots, statement should not be interpretedextensively experienced pilots), that present training programs are

deficient; rather it indicates thee The ship model characteristics healthy attitude of professionals

(ship hydrodynamics, environmental seeking to improve the quality ofhydrodynamics, etc.) are particu- their service.larly critical, and require exten-sive validation/modification by * The generic data base was found topilots prior to training. Pilots be generally well-accepted by theare especially sensitive to the pilots, with specific comments cit-vessel model's handling character- ing the desire for port-specificistics. Furthermore, the pilots training within a generic port datagenerally expect a poor handling base if conducted in an appropriateship in a simulator-based training manner.program, since it represents theextreme set of conditions which In summary, the questionnaire opinionsthey believe are more appropriate correspond very well with the experi-to such training, mental results. The empirical data

unequivocally demonstrate that simula-* Pilots expressed an interest in tor-based training can be effective

having a large number of short for pilots. Pilot observations con-runs, rather than several relative- firm this finding, and support thely long runs. Although this may be potential effectiveness of such train-at variance with their actual ex- ing. Pilot support of simulator-basedperiences at-sea, it provides the training, based on the small samplepilots with an opportunity to ex- size used during this investigation,plore a wide range of conditions was not unequivocal. Rather, it wasand possible responses. highly qualified. Simulator-based

training is supported as having poten-* The length of the training modules tial by pilots, but only when properly

were generally considered to be too used in direct correspondence to localshort. This was expected, as a conditions and issues, and with regard

54.

p --"

IMW 4 W

to the particular pilot/trainee back- training does have potential forgrounds. As one pilot implied, this training pilots, and that it is stillexperimental training program con- in the developmental stage.firmed his belief that simulator-based

55

'r'r

'9 ~--r~t.r - I cw q . r. -r -- 2" 1-. --- 1-- - M

CHAPTER 5

CONCLUS IONS

The major issues of interest to this 2. Simulator-based training is effec-investigation were: (1) the potential tive for addressing Advanced In-effectiveness of simulator-based strumentation and Emergency Ship-training for pilots; (2) the pilot handling piloting issues. Thisskill areas for which simulator-based conclusion is based on the empiri-training is potentially effective; and cal data which showed both training(3) the pilot experience levels for modules to be effective, and on thewhich simulator-based training would opinions of the pilots which mndi-be effective. In addition to address- cated the value and potential ofing these issues, this experiment has these training modules.addressed additional secondary is-sues. Conclusions regarding each of 3. Simulator-based training is ex-these, as developed in the previous pected to be effective for most of

9.Results and Discussions sections, are the additional training areas indi-provided below, cated in this investigation (see

paragraph 2.1, Project Approach).This conclusion is reached on the

1. The findings of this experiment basis of the correspondence betweenstrongly demonstrate the potential (1) the findings of the structuredeffectiveness of simulator-based training analysis that identifiedtraining for pilots. The empirical the areas of piloting skill thatdata showed substantial training appear to have potential for simu-gains for the pilots. This was lator-based training, (2) thesupported by the observation of the strong training effectiveness find-

'Npilots themselves, in which a high ings for the two selected modules,proportion felt that they had (3) the opinions of the pilots sup-gained from this training program, porting the two selected modules,and also felt that simulator-based and (4) the pilots' opinions re-training has potential for improv- garding the potential of the thirding pilot skills. Both the empiri- module for training apprentice and

Kcal data and the pilot opinions are limited experienced pilots. Thenot unequivocal in this regard. strong agreement regarding theseThey show strong support for simu- issues supports the contention thatlator-based training, but do not similar results would likely per-indicate that all such training tamn to the other potential areaswould be either desirable or effec- for simulator-based training iden-tive. Rather, these findings indi- tified in paragraph 2.1.cate the potential for such train-ing if conducted under the right 4. Simulator-based training appears toconditions and conducted appropri- have potential for apprenticeately. pilots, licensed pilots with li-

mited experience, and licensed

57

__~ ~ ~ t i: ~ ..

pilots with extensive experience, training, this was not found to beThe particular aspects of piloting the case. Rather, the apparentlyskills for which such training will negative opinion stems from (1) thebe effective is likely to differ belief that pilot training needssubstantially for each of these are beyond simulator capabilitiesgroups. Both the empirical data (i.e., which they are for certainand the opinions of the pilots in- skills, although other skills candicate differential training effec- be properly addressed via simula-tiveness for different groups of tor-based training); (2) the beliefpilots (i.e., extensive experience that simulator-based trainingversus limited experience). Unfor- should not be used in place of cur-tunately, the empirical data are rent training practices (rather, itinsufficient to identify the most has potential value to augment cur-appropriate- areas of training for rent training programs and hencethe different levels of pilot ex- upgrade piloting skills); and (3)perience. Nevertheless, they do the belief that the application of

"C.support the contention that the simulator-based training will bebenefits to be gained from simula- required of pilots and misdirectedtor-based training will vary for as a result of regulators not beingeach group of pilots with regard to sensitive to particular pilot localthe particular shiphandling skills needs and practices. When address-addressed. ing the potential merits of simula-

tor-based training unencumbered by5. The effectiveness of simulator- these other considerations the

based training is likely to be pilots were generally very positive.heavily dependent on the particularpiloting issues associated with 7. A generic data base for piloteach port, and on the specific training can be effective. Thebackgrounds of the individual empirical data demonstrated the

*pilots. This conclusion has been effectiveness of pilot trainingreached on the basis of the empiri- using a generic data base. Pilotcal data showing differences be- opinion was largely in agreementtween the groups of pilots, obser- with the effectiveness of a generic

Svations of substantial individual data base. However, the particulardifferences within groups during situations, skills, conditions,the experiment, and the opinions of etc., addressed during the trainingthe pilots. For this reason, simu- program should be tailored to thelator-based training cannot be said pilots' local port characteris-to be effective for all pilots in tics. In other words, the geogra-all ports. Rather, the effective- phy need not be the same, but theness of such training will depend issues addressed (e.g., bank ef-

*on the skill areas being addressed, fects) should be those of particu-particular port characteristics, lar relevance in the pilots' port.and the individual pilot/trainee This finding is particularly impor-

*background. tant, since it strongly suggeststhat effective pilot training can

* 6. Pilot opinion of simulator-based be achieved without the substantialtraining is, for the most part, expense of modeling each pilots'positive. Although pilot opinion port. It should also be noted,has been generally thought to be that this finding does not mean tonegative toward simulator-based imply that all piloting skills ad-

58

dressable via simulator-based conclusion would apply equally totraining can be effectively other aspects of simulation fide-achieved using a generic port. lity closely aligned with pilotRather, the generic port may be tasks (e.g., bank interaction withappropriate for training a large ownship).

* proportion of such skills.9. The mechanics of the training pro-

8. The training system characteristics gram, in terms of class sizeare particularly important for the (i.e., four pilots), mixture oftraining of pilots. This impor- extensive and limited experiencedtance appears greater than that for pilots (i.e., usually two in eachthe training of deck officers. The group), the case study/seminartraining program content should be training method used for programsspecifically tailored to meet the involving experienced students,needs of each group of pilots, the coordinator roles provided byThis would include the skills being the pilot consultants and so on,

*addressed, as well as the situa- represent an effective trainingtions and conditions contained in program strategy for the conductthe exercises. A relatively large of piloting training. This con-number of short-duration exercises clusion is based on the careful

*should be employed in the training analysis an~d development workprogram. These should focus on the leading to the training program,extreme situations likely to be the strong empirical results de-encountered by the pilots. They monstrating the effectiveness ofshould provide the pilots with a the training program, and therelatively difficult challenge, and opinions of the pilots showingenable them to practice a variety support for the training program.of techniques that safety and otherconsiderations would not allow try- 10. A governmental requirement for alling at-sea. Finally, the level of pilots to participate in simula-simulation fidelity is very criti- tor-based training at thiis timecal. The pilots were unanimously would likely be ineffective. Thiscritical of the ship hydrodynamics, conclusion is reached on the basiseven though considerable care and of the opinions of the pilots.development had been focused on Although they were generallyproviding an appropriate ship strongly supportive of the poten-model. Similar concern had been tial for simulator-based trainingexperienced at other facilities as a means of improving pilotingtraining pilots (e.g., Europort skills, they were equally stronglypilots trained at Wageningen). The adverse to the concept of suchpilots are extremely sensitive to training being required. Thethe hydrodynamic characteristics of findings of this investigation,the vessel, and would prefer a re- although strongly supportive oflatively poorly handling vessel the potential effectiveness ofthat represents the more difficult simulator-based training forend of the scale. Although similar pilots, also suggest that thecriticism was not placed on the effectiveness can varyenvironmental aspects of the simu- substantially depending on manylation, it is expected that this factors.

59

'S. CHAPTER 6

RECOMMENDATIONS

As a result of the conduct and analy- 2. An investigation should be under-sis of the experimental training pro- taken to identify those pertinentgram described within this report, areas of shiphandling skill forseveral pertinent recommendations have which cost/effective simulator-been identified and discussed below, based training can be provided forIt should be noted that these recoin- pilots. The investigation shouldmendations are the authors' recommen- thoroughly document those skilldations relating to the implementation areas having potential benefit fromand future researc 'h of marine pilot simulator-based training, andsimulator-based training. They do not determine those skill areas fornecessarily reflect the views or in- which simulator-based trainingtentions of either the U.S. Coast would be effective. The finalGuard or the Maritime Administration evaluation of effectiveness shouldwho sponsored this research. be based on a combination of struc-

* tured subjective analysis and em-pirical research. The intent would

1. Government agencies should posi- be to develop a thoroughly docu-tively encourage pilots to avail mented catalog of pilot skill areasthemselves of simulator-based amenable to simulator-based train-training. Such training should not ing. The first step of this cata-be required. As evidenced by this log is the guidelines document de-

*experiment, simulator-based train- veloped by this project addressinging of pilots is in the develop- simulator-based training systemmental stage. It has been examined design for pilots. The experimenthere, and is believed to have po- reported herein conducted a limitedtential for the acquisition, main- analysis of this type, identifyingtenance, an improvement of selected several possible areas of pilotingpiloting skills. It should, there- skill amenable to simulator-basedfore, be considered as a means for training. A much more thoroughaugmenting/upgrading the current investigation directly designed fortraining practices of pilot organi- this purpose is recommnended to in-zations. It is further recommended vestigate and identify those areasthat the appropriate use of simula- for which simulator-based trainingtor-based training be left to the would be cost/effective for pilots.discretion of each pilot organiza-tion. This would enable each or- 3. Several pilot groups should beganization to cost effectively use identified for which simulator-simulator-based training as one me- based training would be particular-

.1dium of a multi-media approach to ly effective, and also acceptablemeet their particular training to the pilots. Appropriate train-needs. ing programs should be set-up on a

61

cost-shared basis between the is recommended that the governmentpilots and the government. The provide technical training assis-purpose of these initial training tance for the development of effec-programs would be twofold: (1) to tive training programs.achieve effective pilot trainingwhere needed; and (2) to encourage 6. An effort should be conducted tothe use of simulator-based training thoroughly develop several repre-on the part of pilots by providing sentative ship hydrodynamicprototype training programs in cer- models. This developmental efforttamn areas. This will help to edu- should include at-sea validation ascate pilots regarding the cost! necessary. Its purpose should beeffectiveness of simulator-based to develop several representativetraining, as well as its accept- models that have met the approvalability. A major problem today of pilot organizations, and whichwith the acceptance of simulator- can be made available to commercialbased training is the lack of dir- training establishments. Theect knowledge by pilots of the hydrodynamic fidelity of the vesselcapabilities and limitations of and the environment are of majorthis training medium. concern to pilots, as evidenced

during this investigation. The4. A high quality generic port data lack of acceptable ship models

base should be developed that will would likely represent a majorcontain representative critical stumbling block to the effectivecharacteristics of ports around the use of simulator-based training bycountry as they pertain to the pilots. Hence, it is in the inter-simulator-based training of speci- est of the industry and maritimefic pilot skill areas. This gener- safety that models be developedic data base should have the char- which meet the approval of pilotacteristics to become a standard- organizations, and hence assist inized training port for pilots, facilitating the simulator-basedThe data base should be developed training of pilots.so as to enable commercial trainingfacilities to transpose its charac- 7. Organizations that intend to useteristics into an appropriate simu- all or parts of the experimental

*lation data base (i.e., visual and pilot training program developed-:radar images, hydrodynamics). A for this investigation should use

catalog of the generic port charac- the data reported herein, andteristics should be crossed-indexed otherwise available, to update

*with the appropriate ports around those training modules. Effective*the country to provide guidance use of this material would require

information concerning potential modification of the modules in ac-%use. This development should in- cordance with the observations of* dude the validation of the generic the pilots and the project staff.

port characteristics. This is considered essential foreffective use of this material.

5. Government agencies should provideassistance to pilot organizations 8. At-sea validation of the findingsto help them in the development and achieved during the Training anduse of simulator-based training Licensing Project should be under-programs. In addition to the fi- taken to provide the bottom-linenancial assistance noted above, it

62

- -.-

information regarding the effec- the data suggest it is a potential-tiveness of simulator-based train- ly strong medium for improvinging. The findings from this expe- shiphandling proficiency. Demon-riment and the earlier investiga- stration of the effectiveness oftions during this project have do- this medium for improving at-seacumented benefits and limitations shiphandling proficiency is a ne-of simulator-based training of cessary pre-requisite for achievingmariners. These findings, however, substantially increased industryhave been confined to the simulator use of this powerful tool and theenvironment. As a result, the use resulting benefits to shiphandling

I of simulator-based training has safety.been growing slowly, even though

63

APPENDIX A

DEBRIEFING QUESTIONNAIRE

The following questionnaire will be administered by one of the research staff.Although responses to many questions are specified, there is a comment spaceavailable after each response. The staff member should encourage comments by thetrainees.

Instructions to Trainee:

In answering the following questions, please do not hesitate to criticize anyaspect of the training program here at CAORF - simulator, seminar coordinator,conditions, etc. Your honest answers, both negative and positive will be of helpto both CAORF and this project. Be specific as possible when asked for comments.

MODULE I: ADVANCED INSTRUMENTATION

1. On a scale of one to ten, ten being excellent, how would you rate the "advancedinstrumentation" segment of the training program?

1 2 3 4 5 6 7 8 9 10

Comments: 7.58 Average

2. Was the amount of time allocated to the use of racons appropriate?

YES NO

Comments:- 8/12 Yes

65

3. Was the amount of time allocated to the use of the rate-of-turn indicatorappropriate?

YES NO

Comments: 10/12 Yes

4. Was the time allocated to the use of the doppler speed log appropriate?

YES NO

Comments: 11/12 Yes

5. Do you feel that you learned something valuable which you can use in yourpilotage port from the "advanced instrumentation" segment of the trainingprogram?

YES NO

Comments: 11/12 Yes

6. What other types of "advanced instrumentation" should be considered for thistype of training?

(Various)

66

7. Do you feel that such training in "advanced instrumentation" would be valuablefor apprentice pilots?

YES NO

Comments: 11/12 Yes

8. Do you feel that periodic optional training in "advanced instrumentation" wouldbe valuable for experienced pilots to keep abreast of the recent developmentsin such bridge equipment?

YES NO

Comments: 11/12 Yes

9. If an "advanced instrumentation" module was offered at a commercial simulatorfacility, what changes should be implemented?

Comments: (Various)

10. Was the simulation of the handling characteristics of the 80,000 dwt tanker

suitable for the "advanced instrumentation" training?

YES NO

Comments: 9/12 Yes

67

SPA -l I

-I I .

4.

9, MODULE II: EMERGENCY SHIPHANDLING

11. On a scale of one to ten, ten being excellent, how would you rate the"emergency shiphandling" segment of the training program?

1 2 3 4 5 6 7 8 9 10

Comments: 6.58 Average4r

12. Do you concur that such training in turning around within confined channels is* a valuable training exercise for developing selected shiphandling skills that

may be appropriate in emergency situations?

YES NO

Comments: 11/12 Yes

13. What other types of exercises do you believe could be appropriate fordeveloping selected shiphandling skills that may be appropriate in emergencysituations?

Comments: (Various)

14. Do you feel that such training in "emergency shiphandling" skills would be

valuable for apprentice pilots?

YES NO

Comments: 11/12 Yes

68

15. Do you feel that periodic optional training in "emergency shiphandling" wouldbe valuable for experienced pilots in order to refresh seldom-used skills?

YES NO

Comments: 9/12 Yes

16. If an "emergency shiphandling" module was offered at a commercial simulatorfacility, what changes should be implemented?

Comments: (Various)

17. Was the simulation of the handling characteristics of the 80,000 dwt tanker

suitable for the "emergency shiphandling" training?

YES NO

Comments: 8/12 Yes

18. Was there adequate time allocated within the training program for refining

your skill in turning around the 80,000 dwt tanker within the confined channel?

YES NO

Comments: 12/12 Yes

69

'V ~~ Pr{~.%'4 ~ Sw X

MODULE III: DECISIONMAKING

19. On a scale of one to ten, ten being excellent, how would you rate the"decisionmaking" segment of the training program?

1 2 3 4 5 6 7 8 9 10

Comments: 6.29 Average

20. Do you feel that you learned something valuable which you can use in yourpilotage port from the "decisionmaking" segment of the training program?

YES NO

Comments: 3/8 Yes

21. What other types of decisionmaking situations do you believe would beappropriate for development/refinement via such simulator training?

Comments: (Various)

22. "o you feel that such "decisionmaking" training would be valuable for

apprentice pilots?

YES NO

Comments: 7/8 Yes

70

23. Do you feel that such "decisionmaking" training would be valuable for licensedpilots with limited experience?

YES NO

Comments: 8/8 Yes

24. Do you feel that such "decisionmaking" training would be valuable as a

refresher for licensed pilots with extensive experience?

YES NO

Comments: 3/8 Yes

,- 25. If a "decisionmaking" module was offered at a commercial simulator facility,what changes should be implemented?

Comments: (Various)

26. Was the simulation of the handling characteristics of the 80,000 dwt tankersuitable for the "decisionmaking" training?

YES NO

Comments: 4/8 Yes

71

GENERAL

27. Do you believe that shiphandling simulators have a role in the trainingprocess of apprentice pilots?

YES NO

Comments: 12/12 Yes

28. Do you believe that shiphandling simulators have a role in the training

process of licensed pilots?

YES NO

Comments: 11/12 Yes

29. Would you recommend that another member of your pilot association attend asimilar simulator-based training program at a commercial training facility ifthe tuition cost was $3000?

YES NO

Comments: 4/12 Yes

72

30. Would you recommend that another member of your pilot association attend asimilar simulator-based training program at a commercial training facility ifthe tuition cost was $1000?

YES NO

Comments: 7/12 Yes

31. What is the maximum duration that a simulator training program could have

without seriously impacting your job and personal schedule?

2 Days 1 Week 2 Weeks Other:

Comments: 6 -- 2 Weeks; 4 -- 1 Week; 1 -- 2 Days

32. Do you now believe that marine pilots can receive beneficial training on ashiphandling simulator within a generic port?

YES NO

Comments: 11/12 Yes

73

33. Did the computer-generated visual scene cause you problems during thisexperimental training program?

YES NO

Comments: 5/12 Yes

34. Should this training program have included exercises under simulated nighttime

conditions?

YES NO

Comments: 5/11 Yes

74

'e

iz A, Z * zu fi L-.-

BIBLIOGRAPHY

Federal Register, Volume 48, No. 19, Thursday, January 27, 1983, p. 3914.

Gynther, J.W., T.J. Hammell and V.M. Pittsley. Guidelines for Marine PilotSimulator-based Training Systems. National Maritime Research Center, KingsPoint, New York, 1984.

Gynther, J.W., T.J. Hammell, J.A. Grasso and V.M. Pittsley. Simulators forMariner Training and Licensing: Guidelines for Deck Officer Training Systems.U.S. Coast Guard and Maritime Administration, Washington, D.C., 1982.

Hammell, T.J., J.W. Gynther, M.E. Gaffney and J.A. Grasso. Simulators forMariner Training and Licensing. Phas? 2: Investigation of Simulator Characteris-tics for Training Senior Mariners. U.S. Coast Guard and Maritime Administration,Washington, D.C., 1981.

Hammell, T.J., J.W. Gynther, J.A. Grasso and D.C. Lentz. Simulators for MarinerTraining and Licensing. Phase 2: Investigation of Simulator-based Training for

. Maritime Cadets. U.S. Coast Guard and Maritime Administration, Washington, D.C.,1981.

Huffner, J.R. Pilotage in Confined Waterways of the United States: A Prelimi-nary Study of Pilot Decisionmaking. Maritime Institute of Technology andGraduate Studies. Linthicum Heights, Maryland, July 1976.

Huffner, J.R. Pilotage in the Port of New York. Maritime Institute of Techno-logy and Graduate Studies. Linthicum Heights, Maryland, July 1978.

Hussem, J., C. DeBoer and P.J. Paymans. "Seven Years Experience with SimulatorTraining of VLCC Pilots in the Netherlands", Proceedings of the Second Interna-tional Conference on Marine Simulation. Compute Aided Operations ResearchFacility, Kings Point, New York, June 1981.

Maritime Transportation Research Board. Proceedings: Symposium on Piloting andVTS Systems. National Academy of Sciences, Washington, D.C., 1980.

Willms, 0. "The Training of Pilots on A Shiphandling Simulator", Proceedings ofthe Second International Conference on Marine Simulation. Computer Aided Opera-tions Research Facility, Kings Point, New York, June 1981.

75

:.~~N %: .

.- I=

FILMED

11-85

~DTICi " -, " , - --, . . .-. . -, " .' " ' " " " . .-. -L , . ' " " "

'.-4 ~~:~~* P. ... Jil lnl-I Jllll~lllli k llI~ i i l

Documents

AID AL EIEEIIEIIEIIIE - dtic.mil 677 experimental evaluation of simulator-based training for i/i marine pilots(u) national maritime research center kings point nv computer aid t j