SHAREit Tutorials Closed - UPC Universitat Politècnica de

1SHARE‐itIntelligentTutorialsTutorials

INDEX_____________________________________________

INTRODUCTION..........................................................................................................................7

1.MOTIVATION ...............................................................................................................8

2.PURPOSEANDSCOPE ................................................................................................10

2.1Keyproblems ........................................................................................................ 10

2.2Tutorials’Goals ..................................................................................................... 11

2.2.1Whatthesystemdoesnotdo......................................................................... 11

3.STATEOFART ............................................................................................................ 13

3.1IntelligentTutoringSystems.................................................................................. 13

3.1.1ExistingITSforcognitivesupport .................................................................... 14

3.2Multi‐AgentSystems............................................................................................. 16

4.SHARE‐it.....................................................................................................................18

4.1ContributionsandObjectives ................................................................................ 18

4.2SHARE‐itMASStructure ........................................................................................ 19

4.2.1PatientAgent.................................................................................................. 20

4.2.2EnvironmentAgent......................................................................................... 21

4.2.3VehicleAgent ................................................................................................. 21

4.2.4HomeAgent ................................................................................................... 21

4.2.5CaregiverAgent.............................................................................................. 22

5.INTELLIGENTTUTORIALS............................................................................................23

5.1Targetpopulation ................................................................................................. 23

5.2FactsandAssumptions.......................................................................................... 25

SPECIFICATION.........................................................................................................................26


6.CONCEPTS..................................................................................................................26

6.1UserProfile ........................................................................................................... 26

6.2ADLs...................................................................................................................... 29

6.3Interactionworkflow............................................................................................. 30

6.4Ontologies ............................................................................................................ 32

7.ACTORS......................................................................................................................34

7.1User ...................................................................................................................... 34

7.2Caregiver............................................................................................................... 34

7.3Otheractors.......................................................................................................... 35

7.4Hierarchyofactors................................................................................................ 35

8.TUTORIALS’PERFORMANCE ......................................................................................37

8.1Scenarios .............................................................................................................. 37

8.1.1Scenario1:Albertotakeshisdrugs ................................................................. 38

8.1.2Scenario2:Pietrogoestothechurch ............................................................. 39

8.1.3Scenario3:Francescacookspasta .................................................................. 40

8.1.4Scenario4:Annadoesthehousework ............................................................ 41

8.1.5Scenario5:Stefanoreceivesthevisitofafriend............................................. 42

8.2Activities’Description ........................................................................................... 43

8.2.1Domestics....................................................................................................... 44

8.2.2Dressing.......................................................................................................... 46

8.2.3Cooking .......................................................................................................... 48

8.2.4Medical .......................................................................................................... 53

8.2.5Grooming ....................................................................................................... 54

9.REQUIREMENTS ......................................................................................................... 57


9.1Actors’interaction ............................................................................................... 57

9.2FunctionalRequirements ...................................................................................... 58

9.3Non‐FunctionalRequirements .............................................................................. 59

9.3.1Look&Feelrequirements ................................................................................ 60

9.3.2Usabilityrequeriments ................................................................................... 60

9.3.3Operationalrequirements .............................................................................. 60

9.3.4SupportandMaintenancerequirements ........................................................ 61

9.4UseCases.............................................................................................................. 62

9.4.1UseCasesDiagrams........................................................................................ 62

9.4.2UseCasesSpecification .................................................................................. 64

10.RISKANALYSIS ......................................................................................................... 67

10.1NewProblems..................................................................................................... 67

10.1.1UsersandCaregiversProblems..................................................................... 67

10.1.2Follow‐upProblems...................................................................................... 67

10.2Risks.................................................................................................................... 68

10.2.1Personnelshortfalls ...................................................................................... 68

10.2.2Developingthewrongfunctions ................................................................... 68

10.2.3Continuousstreamofrequirementschanges ............................................... 68

10.2.4Tableofrisks ................................................................................................ 69

IMPLEMENTATION...................................................................................................................70

11.TUTORIALS’MODULERELATIONS............................................................................ 70

11.1DLAPackage:TutorialList .................................................................................... 72

11.2DLAPackage:SelectedTutorial ............................................................................ 72

11.3DLAPackage:ConfirmationTutorialSteps ............................................................ 73


11.4DLAPackage:ContextualTutorials ....................................................................... 73

12.JADEX.......................................................................................................................74

12.1IntroductiontoJADEX ........................................................................................ 74

12.1.1TheBDIModelofJADEX ............................................................................... 74

12.1.2AgentSpecification....................................................................................... 75

12.2Tutorials’JADEXclasses....................................................................................... 76

12.2.1ListofTutorials ............................................................................................. 76

12.2.2ContextualTutorials ..................................................................................... 77

13.RULE‐BASEDEXPERTSYSTEM ..................................................................................80

13.1JESS..................................................................................................................... 80

13.2KnowledgeBase.................................................................................................. 81

13.3WorkingMemory................................................................................................ 83

14.MEDIARECORDING.................................................................................................. 84

TESTING................................................................................................................................... 86

15.TESTCASES............................................................................................................... 86

15.1TestingCycles...................................................................................................... 86

15.1.1PlanofAction ............................................................................................... 87

15.1.2Experimentation........................................................................................... 88

16.OUTCOMES .............................................................................................................. 89

16.1Tutorials’ExperimentationResults...................................................................... 89

16.1.1Scenario1:Albertotakeshisdrugs .............................................................. 89

16.1.2Scenario2:Pietrogoestothechurch........................................................... 92

16.1.3Scenario3:Francescacookspasta ............................................................... 94

CONCLUSIONS..........................................................................................................................96


17.TEMPORALANDECONOMICANALYSIS ................................................................... 96

17.1TemporalAnalysis ............................................................................................... 96

17.2EconomicAnalysis............................................................................................... 98

18.CONCLUSIONS........................................................................................................ 101

18.1Futurelinesofresearch..................................................................................... 101

APPENDIX............................................................................................................................... 103

A1.GLOSSARY.............................................................................................................. 103

A1.1Medicalterms................................................................................................... 103

A1.2Technicalterms ................................................................................................ 105

A2.STORYBOARDS ...................................................................................................... 107

REFERENCES ........................................................................................................................... 110



INTRODUCTION___________________________________________

First part of this document is intended to give readers a general idea of the project in its

extension.Theintroductionofthedocumentisdedicatedtoexplainthepurposeandgoalsof

the project, followed by a brief introduction of the context were Intelligent Tutorials were

developed:SHARE‐itproject.

Second part of the document, starting from chapter 6, focuses on the specification of the

system, describing some important concepts, establishing a hierarchy of actors, analyzing

requirements and obtainingUse Cases. There is also a section dedicated to planwhich risks

mightariseastheprojectstartsandhowtomitigatethem.

Thirdpartof thedocument, starting fromchapter11,describes the technologiesused in the

implementationof the system,andhow thecodehasbeenprogrammed inorder todevelop

thesystem.

Last part, starting from chapter 15, has been dedicated to explain in depth how the testing

phaseran,whichweretheresultsobtainedandhowdeviationsweretreated.Atemporaland

economic analysis of the project is also included, aswell as a section dedicated to state the

conclusionsandtodrawsomefuturelinesofresearch.


1.MOTIVATION

Nowadays, improvements inthequalityof lifeareraisingsociety’s lifeexpectancy.Therefore,

governments are currently searching for new solutions to cope with the ageing of the

population,whichdemandsalternatives fortakingcareofelderlypeople.Affordableassistive

systemsforpatientsstayingaloneathomeareconsideredtobeasolutiontoinstitutionsand

geriatrics,sothatitisfeasibletotakecareofolderadultsremotely.

A considerable amount of systems have been developed to help people compensate for the

physical and sensory deficits that may accompany aging. Many of these do not rely on

computertechnology,suchasliftchairsorergonomichandles.However,anincreasingnumber

ofdevicesrelyonadvancedcomputer‐basedtechnologies[31].

Therealizationofacompletesystemfortheaidoftheelderlytobeemployedinacomplexand

real environment such as a health‐care institution or a home requires a tight integration of

Assistive Technologies (AT), which can support elderly peoplewith cognitive impairments in

oneormoreofthefollowingways:

(1)byhelpingalleviatethesocialisolationthatmaystemfrommobilityissues,

(2)bydelayingtheirlossofresidualskillsandcognitivedecline,and

(3)byrelievingcaregiversfromtheneedoffulltimecare.

Additionally, institutionalizationhasanenormous financial cost, notonly foreldersand their

caregivers but also for UE governments, which normally pay for nearly 60% of their annual

nursinghomebill [28].Therefore, themajorscaleusesofATrepresentsapotentialsavingof

enormousamountsofmoney.

The main purpose of using Intelligent Tutorials (IT) in the context of AT is to develop an

intelligentandsemi‐autonomoussystemforelderlypeoplewithdisabilitiestoassistthemona

daily routine, so they do not have to entirely depend on a caregiver. This possibility would

increasetheirqualityoflifeand,atthesametime,delaytheirinstitutionalization.

Anothermotivationpointincreatingsuchanassistivetoolliesintheimprovementoftheuser’s

acceptance, mainly in terms of the impact the use of intelligent systems providingmedical


serviceshasonelders.MuchofthereluctancetousethesesystemssimplyarosebecauseAT

devices do not fit naturally into the process of care and, as a result, using them requires

additionaleffort. It is also true,butperhapsdangerous, toascribe someof the reluctance to

use AT upon the technophobia or computer illiteracy of healthcare workers and caregivers.

Therefore, IT have been designed to be perceived beneficial by users, because otherwise,

independentlyofitstruevalue,theywillprobablyberejected.

Issueswhichneedtobeaddressedwhendesigningand implementing ITare:communication,

knowledgerepresentation,human‐machineinteractionandsymbolicreasoning,onlytonamea

few. The philosophy underlying this application is to enhance the quality of Elder care by

employing an infrastructure of expert systemswhich cooperate in order to provide services.

Thisrequiresthedesignofintelligentagentsendowedwithadvancedlearningcapabilities,able

to process symbolic reasoning and easy to interact with humans at a high level. Therefore,

creatingsuchatutoringserviceconsequentlypresentsanimportantchallenge.


2.PURPOSEANDSCOPE

This bachelor thesis has as its main purpose the development of an application to help

individuals suffering fromdementiaorpost‐stroke to carryout indoordailyactivities, so that

theydonothavetoentirelydependonacaregiver.Usershaveavailableapersonalizedrange

of tutorialsaccording to theirphysical andcognitive impairment,whichalso determineshow

theinformationisdisplayedintermsofinteraction.

Futhermore,ITareconceivedasanassistivetooltobeusedonadailybasisfortheworkoutof

userswithdisabilities throughbasicActivitiesofDaily Living (ADLs), so that theworseningof

theirmentalfacultiesanddeclinephysicalcapacitiescanbedelayedintime.

ThemostrelevantkeyissuesrelatedtotheeldercarethatITtriestosettlearedescribedinthe

nextsection.

2.1Keyproblems

Average life expectancy in the European Union is one of the highest in the world and is

continuingtorise[14].In2000, itwas74.7yearsformenand81.1yearsforwomen;in2010,

accordingtoEurostat’sbasescenario,itwillbe79.7and85.1yearsrespectively[30].

Ontheonehand, increasing longevityand increasingsurvival toacuteaccidentsanddiseases

implyanincreasedprevalenceofchronicmorbidityanddisability.Subjectsaffectedbychronic

diseasesoroutcomesofacuteevents,suchasdementiaorstroke,representaheterogenous

category of individuals [15]. Moreover, different conditions can be often combined with

different severity in individuals users, impairing their self‐dependency and worsening their

qualityoflife.

On the other hand, elderly people account for a high percentage of hospitalisation in most

European countries. Social services are placed under considerable pressure since it iswidely

recognised that hospitalization is sometimes not related to the patient’s acute medical

condition, but to the need for rehabilitation or for social reasons, such as the lack of family

supportorabsenceofeconomicalmeans.


Consequently,hospitalizationisnotseenasanappropiatealternativeduetoitscost:thereisan

urgentneedofdelayingusers’institutionalization,facilitatingaccesstospecialistcarefacilities

andusingtechnologicalsolutiontosupporttheirroutinesathome.

2.2Tutorials’Goals

Goals tobepursuedby ITaredividedaccording to two relevantactors:users,whomightbe

abletocarryoutsimpledailyactivitiesathomewiththeassistanceoftutorials,andcaregivers,

whomightberelievedfromtheeffortthatsupposessupportingusersthrougheverybasictask

alongtheday.

Inordertoverifythat ITaresuccessfullyachievingthedefinedgoals, thefollowingobjectives

havetobeattained:

‐ Toincreasethegradeofusers’self‐dependency

‐ Toenhancethecognitivefunctioningandphysicalabilitiesofusers

‐ Topartiallyrelievethecaregiver’seffortinsupportingtheuser

The ambitious goal of providing support for elderswith the use of intelligent systems is not

merely aimed at automating certain tasks. The implementation of IT is motivated by the

necessityofprovidingconstantcareandsupportforpeoplewithdisabilities,providingdecision

making,remindingandwarningfunctionality.Suchcomplextasksnecessarilyinvolveteamsof

technologies,intelligentsensorsandphysician’sknowledge.

2.2.1Whatthesystemdoesnotdo

IT are intented to support certain ADLs1 with the purpose of getting users to be more

independentondailyroutineswhiletheyexercisetheircognitiveabilitiessuchascoordination

1Refertosection6.2fordetailedexplanationoftheADLs


andattention.Collaboration fromcaregivers is required inorder to fulfill the initial and final

conditionsandsolveanyunexpectedproblem.

The range of tutorials designed do not contemplate those involving the user’ mobility on

activitiessuchaswalkingaroundthehouseortransferringfrombedstochairs.ITnotintended

neithertosubstitutecaregiversnortobeatoolforrehabilitation,althoughitcanhelpfulforthe

users’functionalenhancement.


3.STATEOFART

ITarebasedontwodifferentsystems:

• theconceptualpatterncorrespondtothe IntelligentTutoringSystems(ITS),whichare

based on the learner model, a model constructed from observation of interactions

betweenalearnerandalearningsystemorinstructionalenvironment[27].

• the technology used for IT implementation as an intelligent service is a Multi‐Agent

System(MAS),whereseveralagentsconsideredtobeautonomousentitiesinteractwith

eachotherinordertoshareacommongoalopursuetheirowninterests.

Following, both systems are introduced as well as some similar existing systems to be a

referenceintheAIfieldregardingtechnologiesforthecognitivesupport.

3.1IntelligentTutoringSystems

Many traditional instructional tutorials present learners with facts and concepts. These

methods are effective in exposing people to large amounts of information and testing their

recall. By contrast, other systems use simulations and other highly interactive learning

environmentsthatrequirepeopletoapplytheirskills[8].

ITStypicallyrelyonthreetypesofknowledge,organizedintoseparatesoftwaremodules[9]:

‐ Theexpertmodule represents subjectmatter expertise and provides the systemwith

knowledgeofwhatisbeingtaught.Itisacomputerrepresentationofadomainexpert's

subjectmatterknowledge and problem‐solving ability. Regarding IT, expertmodule is

representedbythemedicalknowledge,speciallyintermsofguidelinesfortheproper

designandspecification.

‐ The learner module represents what the user does and does not know letting the

system know who it is teaching. It evaluates each user's performance to determine

his/herknowledge,perceptualabilities,andreasoningskills.


‐ The instructormoduleenablesthesystemtoknowhowtoteachaccordingtotheuser

profile,byencodinginstructionalstrategiesusedviatheITSuserinterface.

Plus,ITSincludeasubsystemforthe interfacemodule,whichprovidesthemeansforthefinal

usertointeract,usuallythroughaGraphicalUserInterface(GUI)andsometimesthrougharich

simulationofthetaskdomaintheuserisbeingtoldtodo.

3.1.1ExistingITSforcognitivesupport

COACH (CognitiveOrthosis forAssisting aCtivities atHome). A very similar system to IT in

termsoffunctionality,uniquelyaimedatusersaffectedbydementiaandbasedonprompting.

COACHisaprototypeofanintelligent,supportiveenvironmentbeingdevelopedtohelppeople

withdementiacompleteADLswith lessdependenceonacaregiver.Thisworkrepresentsone

of first clinically tested supportive devices to use artificial intelligence techniques. It uses a

personal(desktop)computerandasinglevideocameratounobtrusivelytrackauserduringan

ADL. By sophisticated estimation of a user’s progress through a task using Bayesian filtering

techniques,COACHprovidespre‐recorded(visualorvideo)cuestotheuserwhennecessary.

Thissystemisdividedindifferentmodules,eachofwhichhasadifferentandessentialtaskslike

identifying the users position, deciding to take certain steps depending on the state, or

promptingsomeoftheavailableactions[21].

COACHusesanubiquitousmodelingtechniquethatcanadapttousersovertime:thePOMDP

model.Theideaistohaveasinglemodelandlearningtechniquethatcanbeeasilyappliedto

different tasks,without theneed to re‐engineerthemodel. Thekey strengthsof thePOMDP

model are that it is able to deal with uncertainty, it is easy to specify, it can be applied to

differenttaskswith littlemodification,and it isableto learnandadapttochangingtasksand

situations[16].


MAPS (Memory Aiding Prompting System). Another system which also uses tutoring

techniquesaswellasAItechnologiesisMAPS,whichbreaksdownataskintoconstituentparts

andusesimagesandverbalinstructionsforpromptingeachstep.TheMAPSprojecthasbeen

designed using the perspective that high functioning assistive technology has two equally

importantuserinterfaces:thepersonwithcognitivedisabilitiesandthecaregiver.

Computationally enhanced prompting systems can provide a bridge to independence.

Unacceptableabandonmentratesforhighfunctioningassistivetechnologyarepartiallydriven

bydifficultiesin(re)configurationofthedevice;typicallybyacaregiverwho isnotnecessarily

technicallysavvy[4].MAPSprovidesanenvironmentinwhichcaregiverscancreatescriptsthat

canbeusedbypeoplewithcognitivedisabilities,sothatthesystemismanuallypersonalized

forthespecificneedsoftheuser.

This system is strongly based on hardware: a PDAprovides error correction functionality via

dynamicscriptingand‘panicbutton’functionality,whileaPCbasedapplicationprovidestools

for script creation,modificationand sharingwithotherusers. Severaldifferent interfacesare

displayeddependingontheuser(patient,installer,caregiver).

LISTEN’s Reading Tutor.This device is a good example of an ITS aimed at userswith similar

capabilitiestoIT’stargetpopulation:userswithlowreadingabilityandlowcognitiveskills.

TheReadingTutorinterveneswhenthereadermakesmistakes,getsstuck,clicksforhelp,oris

likelytoencounterdifficulty.Itmayintervenebeforeastory(e.g.topreviewnewvocabulary),

beforeasentence(e.g.toinsertanautomaticallygeneratedcomprehensionquestion),duringa

sentence(e.g. togivehelponaword),and/orafterastory(e.g. torevieworpost‐testwords

from the story). The system respondswith assistancemodeled after expert reading teachers

butadaptedtothetechnology’scapabilitiesandlimitations[17].

Itsinterfaceisdesignedtokeeptheinteractionsimpleandtofocusattentiononthetext.Users

pickwhichstorytoreadnext,whicharecomposedofafewtypesofsteps.


LISTEN’salsopursue theupdateof theuser cognitive state,continuallyestimating theusers’

readingabilitybasedonhowlongtheyhesitatebeforeeachwordtheyread.Thesystemlater

usesthisdatatoadjustthelevelofstoriesitchoosesonitsturn.

3.2Multi‐AgentSystems

Agent‐basedsystemstechnologyisparticularlyattractiveforcreatingsoftwarethatoperatesin

environmentsthataredistributed,suchaseldercare institutionsandusers’home.AsATget

morematureandaddressincreasinglycomplexapplications,theneedforsystemsconsistingof

multipleagentsthathaveapeer‐to‐peercommunicationisbecomingapparent[14].

A MAS is a system composed of multiple interacting software entities called Agents which

continouslyperformmultiplefunctions:perceptionofdynamicconditionsintheenvironment,

reasoningtointerpretperceptions,anddeterminationoftheactionstosolveproblems[32].

Figure1.MultiplefunctioningAgent

Agentscollaborateamongthemselvesinteractingthroughasetofreciprocalactionstoachieve

thepursuedobjectivesusingparticularmechanismsandprotocols.


An agent dynamically constructs explicit control plans to guide its choices among situation‐

triggered behaviors. Each agent composing a MAS may have its own task and role, and

sometimesastrongernotionisadded(Beliefs,Desires,Intentions)[26].AgentsandMulti‐Agent

Systemsareusedasametaphortomodelcomplexdistributedprocesses.

Adistributedsystemisconsideredtobeacollectionofindependentsystemsthatappeartothe

usersasasinglesystem.Bothhardwareandsoftwarecomponents,whicharedistributedand

connectedbynetwork,coordinateandcommunicateitsactionsthroughmessagesinorderto

reachagoal.Communicationisestablishedbyaaclient‐serverprotocol.

Bothapproachessharethenotionof‘distributedness’.Theareaofdistributedagentcomputing

istheareainwhichtheseapproachesintersect,enablingandsupportinglarge‐scale,secureand

heterogeneousprocesses.


4.SHARE‐it

SHARE‐it(SupportedHumanAutonomyforRecoveryandEnhancementofcognitiveandmotor

abilitiesusingInformationTechnologies)projectaddressestheissuesofenhancingthequality

oflifeandindependenceofelderlypeoplewithcognitiveandphysicaldysfunctions.Thegoalof

SHARE‐itistodevelopascalable,adaptivesystemofadd‐onstosensorandassistivetechnology

sothattheycanbemodularlyintegratedintoanintelligenthomeenvironmenttoenhancethe

individual’s autonomy. The system is designed to inform and assist the user and caregivers

throughmonitoringandmobilityhelp.

4.1ContributionsandObjectives

SHARE‐it is composedbyconjunctionofdifferentelements, suchas sensornetworks, robotic

platformsoraMASinchargeofprovidingassistiveservices.ITisoneofthesemultipleservices

that SHARE‐it project offers, guiding users on how to carry out daily activities such as

housekeepingorcookingaccordingtothegradeoftheirdisability.

Moreover, thisproject isplannedtocontributetothedevelopmentofthenextgenerationof

assistive devices for older persons or people with disabilities, so that they can be self‐

dependentas longaspossible.SHARE‐it is focusedonadd‐onstobecompatiblewithexisting

technologiesandtoachieveaneasierintegrationintoexistingsystems.

Inthiscontext,SHARE‐itmakessignificantcontributionstofundamental,long‐termresearchin

thefollowingareas[22]:

• C1. Sensor‐based environment perception, knowledge acquisition and representation,

high‐levelreasoningandgoalseekingbehaviorsinarealworld.

• C2.Verifyingsoftwareadaptationtohumanwithspecialneeds:bothatdesignandrun‐

timetoestablishsafety,regulatoryandsecurityrequirements.


• C3. Incorporating shared autonomy: ensuring that software components can be

designedtooperateinagivenintelligentambianceandadapttopossiblechangesboth

intheneedsoftheuserorintheenvironment.

TheobjectivesofSHARE‐itare:

• O1. To explore the benefits of the concept of situated intelligence to build elements

(add‐ons)thatwillenhancetheautonomyofthetargetusergroup intheirdailylifein

theirpreferredenvironment.

• O2.Toinvestigateandimplementinnovativeformsofsharedautonomy.

• O3. To build appropriate add‐ons to standardized technologies to provide ubiquitous

sensing,computationandassistance.

• O4.Tobuildadaptiveinterfacesforthetargetgroup.

• O5. To target the various human‐delivered assistance and caretaking services as

effectivelyaspossible.

4.2SHARE‐itMASStructure

Oneof the software components of SHARE‐it is theMAS,which in charge of the knowledge

engineering,retrievinginformationfromtheenvironment,reasoningtointerpretperceptions,

andactingtopursuefixedgoals.AgentsgettheirinputdatafromotherelementswithinSHARE‐

it,suchthesensornetwork,themedicalhistorystoredintheserver,orthemobilityplatforms.

TheMASiscomposedby5differentsagentsinteractingamongthemselvesinordertoachieve

thedesiredgoals:PatientAgent,EnvironmentAgent,VehicleAgent,HomeAgentandCaregiver

Agent,eachofthemprovidingdifferentservices,asitisillustratedintheMASschemabelow.


Figure2.DistributionofSHARE‐itAgents

4.2.1PatientAgent

ThePatientAgentencasesmostofthecognitivesupportservicesorientedtotheusersuchas

IT(theserviceisalsoknownasADLTutor).Aninstantiationofthisagentshouldprovideallthe

availableandpermittedservicestoeachusersuchassecurity,mobility,monitoringandhelp.

Those services will be offered depending on the user disability, so that not only a complex

profilingworkhastobedonewiththeuser,butalsoadefinitionoftheservicestobelaunched

andhowtoadaptthemtotheuser.

ThePatientAgentisinchargeofallthefunctionsrelatedtoADLs,suchaskeepingatrackofthe

currentADLsbeingperformedby theuser in order todetermine if something importanthas


beenforgotten.Thisagentisalsoresponsibleformanagingremindersandrequestinghelpfrom

caregiversondemandorunderasetofparticularcircumstances.

4.2.2EnvironmentAgent

TheEnvironmentAgentgathers informationfromtheenvironmentandgives it semanticmeaning

relatingittoontologicalconcepts(e.g.thelocation)usedbytheagentsystem[33].

Intermsoffunctionalities,thisagentisinchargeofthereadingfromthesensornetwork,filtering

data with a given criteria, managing positioning information and actuator status, and detecting

possibleemergencysituations.

ThisagentrepresentsanimportantinputfortheITservice,sinceitisresponsibleofinformingnot

onlyaboutthecurrentpositionoftheuserbutalsoaboutthetimeofchanging.

4.2.3VehicleAgent

The Vehicle Agent is intended to keep track of different status indicators (stability, battery

change level, hardware failures, etc). This agent tracks users’ navigation efficiency, and

dependingontheirprofile,theagentmakedecisionsaboutthedegreeofcontroluserswould

haveonthenavigation[20].

ThemostremarkableroloftheVehicleAgentisthefalldetection,whichretrievesinformation

comingfromthesensorsinthemobileplatform,triggeringthesendingofanalarmviaSMSto

thecaregiverincaseofafalldetection[33].

4.2.4HomeAgent

TheHomeAgent is in chargeof retrieving the knowledgeextractedbyotherMASagents, storing

the information and delivering it when requested. Therefore, the Home Agent is not directly

connected to the sensors or interfaces, but it is closely connected to other agents related to

mobilityinordertorespondtodatabasequeries[33].


The Home Agent has an important role for the IT service, since it stores the user profile and

managesanyupdatemadebydoctorsorcaregivers.

4.2.5CaregiverAgent

TheCaregiverAgentisinchargeofdeliveringinformationaboutthestatusofeldersandtheir

environment to caregivers. Keeping in mind that the informal caregivers are not necessarily

technicalormedicalexperts,theinformationisshowninaverysimpleandintuitivemanner[33].

One of the aspects to be carefully considered is the privacy when access to data. This agent

guaranteesthatonlyauthorizedsubjectsaccesstheinformationrelatedtothoseuserstheyarein

chargeof.


5.INTELLIGENTTUTORIALS

ITareconceivedasanassistivetooltoexerciseuserswithdisabilitiesonadailybasisthrough

theperformanceofindoorADLs.ItcomposesauserserviceofthePatientAgent,integratedat

the same time within SHARE‐it MAS. Tutorials are displayed to the user through the touch

screen assembled on robotic assistive devices. Information inputs are retrieved from the

environmentsensors (EnvironmentAgent)andtheroboticassistivedevices (VehicleAgent) in

ordertotriggertutorials’activationdependingontheuserprofile(HomeAgent).

EachoftheselectedADLstobeperformedwithITsupportisdividedintoseveralsimplesteps

whereinstructionsaretextuallydisplayedaswellasaudiorecorded.Avideoaccompaniesthe

instruction given, guiding users on how to perform the task. Only when the current step is

completed,userscanmoveforwardtothenextstep.Usersarealsoabletorequesthelpincase

ofneed.

OneofthemainpurposesofITistodelayusers’institutionalization,makingthemeasytostay

at home with a reasonably good level of comfort. It is frequent that elderly people with

cognitiveandphysicaldysfunctionsgetworseas timegoesbybecauseof theage factorand

personalconditions.TheIT’sintelligentlayerisresponsiblefortheselectionofthesuitableset

of tutorials each user is able to perform according to his/her disability status, and the

adaptationoftheircontentwheneverthesystemisinformedaboutchangesintheuserprofile.

Dependingonthecertaincontext(e.g.thesystemdetectsthattheuserisinthekitchenviaa

networkofsensors),theserviceselectsasubsetofcontextualtutorialstheusercanperformin

thatlocationaccordingtohis/herprofile.

5.1Targetpopulation

Fromtheservice’spointofview,peoplewithdisabilitiesareviewedashavingcertainpotential

ofindependentliving,eachpersonhavinganindividualneedofassistance.Naturally,thelevel

of independence strongly depends on the individual level of disability. IT are conceived to

support the independence level each user presents, offering different types and degrees of


assistance.Thislevelofgranularityinsupportingusersisseenasanimprovementinthereview

oftheuser’scurrenthealthstatus,which isakeypieceforenhancingtheirskills inshortand

longterms.

Different groups of elders are characterized by large variations in well‐being, disability, and

health care needs. IT target population has to be put into groups, so that a number of

individuals as large as possible can achieve benefits as effectively as possible. Therefore,

differentgroupsofdisabilitiesareclassifiedasbasicusergroupssharingcertaincharacteristics

[15]. Thisclassificationwouldbeuseful todefinetutorials’ typeof interactionwithend‐users

accordingtotheirprofile.

Figure3.Disabilities’featuresaffectingITtargetpopulation

The typical user defined by the medical team is an elderly patient, with cognitive and/or

physical disabilities on mild or moderate degree, co‐morbid conditions, functional loss from

multipledisabilities,leadingtoimpairedself‐dependency.

Inordertogetabetterdefinitionofthetargetpopulation,mostcommonconditionswithhigh

prevalenceareconsidered2.

2RefertoannexA1.1forafurtherextensiveexplanationoftheconsidereddisabilities


5.2FactsandAssumptions

ThemainfactsandassumptionstobeconsideredbeforethespecificationanddesignofITare

listed below. They provide a good vision of the IT service’s boundaries. Functional and non‐

functionalrequirementsmustperfectlyfitallthefollowingpoints.

• ADLsaretheconceptinwhichITarebased.Theycanbelooselydefinedasaclusterof

simpleandexplicitactionsthataredoneforapurpose.

• Activitiestocarryoutbyusinganyofthetutorialsshouldnotbelongtimeconsumingso

theuserdoesnotgetdistractedorlost.

• Itistakenforgrantedthattheinitialconditionsfortheactivitiestostartarefulfilledby

thecorrespondingcaregivers.

• Actions to carry out individually to complete an activity are adjusted to the current

physicalandcognitiveconditionsofusersatanytime.

• The responsiveness of IT can be characterized by the action‐reaction principle:

whenevertheuseriswillingtomoveforwardtothenextstep,thesystemwillassume

s/hehasalreadycompletedcurrentstep.

• Usershavethechancetostartovercertainstepssothattheactionstakenbeforedonot

disruptonthetutorial’sperformance.

• Althoughtherangeofready‐to‐servetutorialsdependsonthecontextwheretheuseris

located,alltheothertutorialsareavailableatanytime.

• Doctors and geriatricians together with caregivers are responsible for updating the

currentcognitiveandphysicialconditionofusersontheircorrespondingprofile.

• ITarelikelytobeextended,sothatotherrequiredtutorialsdevelopinthefuturewillbe

compatibleandeasytointegratewihthecurrentsystem.

• Italianisthelanguageusedfortheinteractionbetweenthesystemandend‐users.The

resultingsystemwouldbeflexibleenoughtoacceptotherlanguagesconfiguration.


SPECIFICATION___________________________________________

6.CONCEPTS

The following section is dedicated to make an introduction of several concepts with an

importantroleonthedefinitionofITservice:userprofile,whichfeaturesgroupsofend‐usersin

ordertocontextualizeandselectthemostconvenienttutorialsaccordingtotheirhealthstatus;

ADLs, activities in which procedural tutorials are based; ontologies, entities which get its

semanticmeaningfromthe input/output informationthatagentshandle;and IT’s interaction

workflow,whichisakeyfactorforthesuccessoftheservice,asitdefineshowend‐usersaccess

toavailablefunctionalitieswheninteractingwiththesystem.

6.1UserProfile

The medical team supervising the specification of IT has considered two type of chronic

disabilities concerning its target population: cognitive impairment and mixed impairment.

Followingdescriptionhavebeendetailedbydoctorsandgeriatricians.

Disability deriving from cognitive impairment: Dementia. Users presenting a mild grade of

dementiamight present progressive difficulties but can follow established routines at home,

whileuserswithamoderategradeofdementiapresentasignificantimpairedabilitytoperform

simpledailyactivitiessuchasbathingordressingastheybecomedisorientedeveninfamiliar

surroundings.

Typically, intellectual and other cognitive functions decline inexorably over 2 to 10 years.

Depressionaffectsupto40%ofpatientswithdementia,usuallyonamildormoderatestate,

andmaycausevegetativesymptomssuchaswithdrawal,weightlossorinsomnia.Alzheimeris

aformofdementia[7].


Disabilityderivingfrommixedimpairment:Post‐stroke.Strokeisaleadingcauseofdisability.

Among survivors, 40%have amoderate functional impairment and 15% to 30% are severely

disabled.

After a stroke,many people cannot do activities that require fine coordination, so that they

cannot move their hands precisely or coordinate the movement of their hands and eyes.

Patientsmay learn newways to do activities such as fastening clothes, opening and closing

containers, and getting objects that are too far to reach. Some patients are partially or

completelyblindinoneorbotheyes.Speechtherapistsmightbenecessarytoretrainpatients

to use those muscles involved in speaking and breathing and also help with swallowing

problems[15].

Inadditiontochronicdegenerativeconditionssuchasdementiaorpost‐stroke,standardusers

might be affected by general declines inmotor coordination, visual and auditory acuteness,

muscleandbonestrength,mobility,andsensoryperceptionsofstimuli(heatandcold)[7].

MentalFunctions:

Consciousness,Orientation(time,place,person),Intellectual(incl.Retardation,dementia),Energyanddrivefunctions,Sleep,Attention,Memory,Emotionalfunctions,Perceptualfunctions,Higherlevelcognitivefunctions,Language

NeuromuscularMovementRelatedFunctions:

Mobilityjoint,Musclepowerandtone,Involuntarymovements

SensoryFunctionsandPain:

Seeing,Hearing,Vestibular(incl.Balancefunctions),Pain

SpeechFunctions: Voice

Figure4.GeneraldeclinestobeconsideredfortheIT’suserprofile

Afterseveraldiscussionswiththemedicalteam,thefinaluserprofilewasspecifiedtakinginto

consideration thepreviousqualifiers listed,but summarizing its contentto themost relevant

anddescriptivefeatures.


The grade of affection a user can present is expressed using quantifiers: 0 ‐ none, 1 ‐ mild

degree, 2 ‐ moderate degree and 3 ‐ severe degree. However, since users suffering from a

severedegreeofimpairmentarenotconsideredastargetpopulation,thisgradeisnotusedto

evaluate any of the user profile’s qualifiers except from sound impairment, as tutorials are

supportedbysoundandtextatthesametime.

Neglectaffectionreferstoaconditionwhichreducesaperson’sabilitytolook,listenormake

movementsinonehalfoftheirenvironment.Therefore,itsquantifiersarerangeddifferently:0

‐none,1‐left,2–right.

Aphasiaaffection refers toa languagedisorderwhich ispresentedwhen there isdifficulty in

usingorunderstandingspokenandwrittenlanguage.Dualquantifiersareusedinthiscase:0–

usernotaffected,1–useraffected.

Following,theuserprofilespecificationispresented.Thequalifiersincludedarediverseenough

tofacilitatethedivisionofthetargetpopulationintogroupsofuserssharingsimilaritiesintheir

profile.

USERPROFILEQUALIFIERS QUANTIFIERS

Visualimpairment 0‐none1‐milddegree2‐moderatedegree

Soundimpairment 0‐none1‐milddegree2‐moderatedegree3‐severedegree

Finecoordination 0‐none1‐milddegree2‐moderatedegree

Aphasia 0‐no1–yes

Neglect 0‐none1‐leftside2‐rightside

Dementia 0‐none1‐milddegree2‐moderatedegree


6.2ADLs

The selected activities listed below refer to daily tasks whose procedural tutorial might be

displayedtoeldersdependingonthetypeofdisabilitytheypresent.Thecomplexityofevery

taskalsodefines thenumberof steps its tutorial is composedby (e.g. cookingpasta tutorial

complexitydiffersdependingonthegradeofaphasia,dementiaandneglectauserpresents).

This classification has been specifiedwith the support of themedical team, considering the

mostcommonADLsthatatypicaluserperforms,andcategorizingthemaccordingtolocations.

Each of the tutorials has been classified, at the same time, according to the grade of

impairmentauserispermittedtohaveforitsperformance.

The process of featuring tutorials according to the type of impairment considered (cognitive

andmixed)isamustinordertoimplementtheiradaptabilitytothecurrentuserprofile.

CATEGORY TUTORIALS COGNITIVE MIXED

Washingthedishes Mild MildDOMESTICS

Washingclothesbyhand Mild Mild

Goingoutdoors Mild MildDRESSING

Puttingshoeson Moderate Moderate

Preparingcoffee Mild Mild

Cookingpasta Mild/Moderate Mild/Moderate

Makingasandwich Moderate ModerateCOOKING

Preparingcoldsoup Mild Mild

MEDICAL Takingdrugs Moderate Moderate

Morninggrooming Mild MildGROOMING

Brushingteeth Moderate Moderate

Figure5.Tutorials’classificationaccordingtogradesofcognitive/mixedimpairment


It is important to remark that“Putting shoeson” and“Brushing teeth” tutorialsareaimedat

usersonoccupationtherapy.Therefore,thesetutorialsarenotaimedatuserswithmilddegree

ofdisabilities.

6.3Interactionworkflow

Due to the type of disabilities presented by IT’s target population, most of them implying

degenerationoftheusers’cognitiveabilities, interactionworkflowdefiningtutorials’ lifecycle

might change from a standard level of comprehension to an easy level where constant

interactionispursued.

Oneofthemorerelevantaspectsofthesystemconceptionistheunderstandingprocessofthe

tutorials’ instructions:it islikelythattheusermightgetdistractedorsimplyforgetwhats/he

wastoldtodo.Elderlypopulationarelikelytoretainmoreinformationwhentheyinteractas

opposed to when they are just being told facts. Therefore, the system pursues a constant

interactionandrevisiononwhattheuserisdoingsos/hecanbeguidedeasily.Thisindividual

comprehension is guaranteed by using confirmation messages every time the user goes

forwardtothefollowingstep.

There are also certain steps considered to be dangerous, as they involve risky tasks such as

turning on the electric cooker. The realization of these steps would be time‐controlled for

safetyreasons.

Regarding the functionality of the service, certain options have to be within reach of users

when performing a tutorial: requesting help at any step or exiting the tutorial are options

availablefortheuser,whoshouldconfirmtheactionpriorto itsexecution.Thepossibilityof

sendinganalertmightbealsoavailableatthefirststep,whentheuserchecksifalltheitems

arearranged.Ifanyofthemismissing,analertissenttothecaregiverandthetutorialcanbe

cancelled. Both functionalities giving help and sending alerts are implemented as another

servicewithinSHARE‐it,whoseinteractionwithITismanagedbytheMAS.


STEP 1.

Check items

Confirm

message

Message: alert reported

to the caregiver

STEP N

Confirm

HELP

Confirm

EXIT

Confirm

message

EXIT

STEP N+1

YES

NO

NO

YES

onwards

onwards

click EXIT click HELP

backwards

Message: help

requested to caregiver

Confirm

ALERT

click ALERT

YES

YES

NO

NO

YES

NO

Figure6.InteractionworkflowforaproperITservice’susability

Final users might participate on the usability validation. It is mainly focused on testing the

human‐systeminteractionwhenfollowingtheworkflowinFigure6,detectingfunctionalerrors

andnewvaluestobeaddedinordertoimprovethesystem.ItisexpectedusersmightuseITon

a daily basis, nomatter which grade of impairment they are affected by, so that it is really

importanttoreceivefeedbackfromdoctorsandgeriatricianspriortoitsdeploymentinorderto

consideranypossibledeviation.


6.4Ontologies

Anontologyistherepresentationofentities,ideasandevents,alongwiththeirpropertiesand

relations,accordingtoasystemofcategories.Thespecialmotivationforusinganontology is

thatagentsaredealingwithreal‐worldentitieswithdifferentgradesofconnection.Therefore,

avocabulary‐likemechanismisneededtoreferandmanagethem[20].

Following,theontologiesdevelopedpriortothe ITspecificationbythetechnicalandmedical

teamareillustrated.Bothontologiesreferentitiesusedtoimplementthecontextualizationand

selectionoftutorialstobeperformedbyusers.

• ADLtracking.Theontology’sconceptsareusedtodiscriminateasetofpossibleactivities

thattheusercouldbeperforming.

Figure7.ADLOntology

• ADL performance location. It is also highly important to knowwhere an activity is being

performedinordertooffercontextualtutorials.


Figure8.ADL’sLocationOntology


7.ACTORS

Actors are those subjects/objects which take part in a sequence of activities as part of a

dialoguewiththeITservicetoachievesomegoal.Actorsmaybeendusers,othersystems,or

hardwaredevicesexistingoutsidethesystem.

7.1User

The typical end‐user has been already defined within the IT’s target population. Users

interactingwiththesystemmighthavemildormoderatedisabilitiesaffectingtheirreasoning

capabilities and physical abilities, but are expected to have a low loss of visual and sound

acuteness.

Aprevious introductiontothesystemmustbecarriedoutbycaregivers inordertofacilitate

thehuman‐systeminteraction,testthesystemusability,andstregthentheuserself‐confidence

whenusingnewtechnologiesforthefirsttime.

Actorname:User

Role: Principalactoron the interactionwith the sytemwhosedifficultiesonperformingdaily

activitiesareintentedtobelessened.

Subjectmatterexperience:Novice.Usersinteractswiththesystemforthefirsttime.

Technologicalexperience:Novice.Non‐existentknowledgeoftechnologybasics.

7.2Caregiver

Therearetwotypesofcaregivers:thosewhoareusers’relativesandthosewhoarespecialitsts

paidbyinstitutionstogivecareservicestousers.

Bothtypeofcaregiversareresponsible forthecoverofdailyneedsusersmighthaveandfor

givingsupportontasksuserscanhardlycarryoutbythemselves.Thesetasksdifferfromthose

tobeperformedeverydaysuchasdressingortoileting,andthosethatare importantforthe

independentlivingofusersinthecommunity.


IntheapplicationofIT,caregivershaveanimportantroleastheyareresponsibleforfulfilling

certainrequirementsinordertoassuretheproperrunningoftutorials.Forinstance,theymust

gatherallthecookingtoolsonthekitchencounterfortheusertoperformthecookingtask,or

they must prepare the right drug dose for the user to take. Additionally, caregivers are

responsible for introducinguserstothe ITapplication,sotheygettoknowhowitworksand

startinteractingwiththesystem.

Actorname:Caregiver

Role: Personal assistant who is responsible for covering clinical needs of individuals with

chronicdiseases,assuringtheircomfortathomeandtrackingtheirrehabilitation.

Subjectmatterexperience:Advanced.Caregiversalreadyhaveastrongexpertiseonassisting

userswhopresentwidelydifferentandheterogenousfunctionalprofiles.

Technologicalexperience:Novice.Non‐existentknowledgeoftechnologybasicsisdemanded.

Caregiversdonotneedtointeractwiththesystemdirectly.

7.3Otheractors

System administrator. In charge of the platform maintenance and the system consistency

track.

Healthprofessionals.Responsibleforthesystemadjustmenttotheusers’requeriments.They

areinchargeofmonitoringusersinordertomakeanevalutionreliableandefficientenoughto

measure their abilities and cognitive functions such as judgment, language, orientation,

calculation,memoryorplanning.

7.4Hierarchyofactors

Following,apriorityisgiventoeachcategoryofactorsinordertomarkitsimportance.

Key Actors. User and Caregiver. Both users are essential for the success of IT as they could

determineifthesystemachievedthedesiredgoals.


SecondaryActors.Thereareseveralsecondaryactorswhosecontributiontotheproperwork

ofITcannotbemissed:

‐ SystemAdministrators,whosetaskisthemaintanceandupdateofthesystem.

‐ Health Professionals, mainly doctors and geriatricians, who are in charge of deciding

about the needs of the users according to their state, health condition and physical

limitation.


8.TUTORIALS’PERFORMANCE

Nextsectionsare intendedtodescribe indetailwhich is thecontextwhereITare conceived,

definingthosetasksanend‐usercanperform,whicharetheconditionstodoso,andwherethe

taskshavetotakeplaceinordertobeproperlycompleted.

8.1Scenarios

Scenariosarenarrativedescriptionsoftypical interactionsbetweenthesystemandtheusers.

Typically,itpresentsasituationwhereactorsandITservice,amongotherservicesofSHARE‐it,

takepart.Themaincharacterisatypicaluseraffectedbyseveraloftheconditionsinacertain

gradethatallowhim/hertoperformITwithoutthecaregiversupport.

Methodologyusedfordescribingeachscenarioincludesthefollowingpoints:

Purposeofthescenario:

describeswhichaspectsofthesystemanditsinteractionswiththeuseraretreatedonthescenario.

Userdescription: a description of the typical user for the scenario. The actorrepresentsanindividualofadefinedtargetgroup.

Narrativescenario: describesanexampleof interactionbetweentheuserandthesystem.

Structureofthescenario:

areferencetotheuserprofileandtothosetasksend‐userscancarryoutwithinthecurrentscenario.

RoleoftheITservice: describes the interactions between IT and the technologySHARE‐ituses.

Roleofcommunications:

describeswhichelementsof theSHARE‐it technologyneed tobeavailableandactiveinthescenario.

Thescenariosfollowing introducedhavebeendescribed inthecontextofSHARE‐itprojectby

themedicalandtechnicalteam,andhavebeenusedduringtheexperimentationphaseinorder

totestITservice.


8.1.1Scenario1:Albertotakeshisdrugs

Purposeofthescenario:thisscenariohighlightshowusefulthemedicaltutorial istosupport

users’therapy.

UserDescription:Albertoisa78yearsoldwholiveswithhisdaughter.Sheworksmostofthe

day. He has been diagnosed with a very early Alzheimer Disease, causing a mild cognitive

disability. His major impairments are related to recent memory and to executive functions

(planningthesequenceofactions).

NarrativeScenario:Albertohastotakesomepillsaspartofhisdrugtherapyafterbreakfast,in

the middle of the afternoon, and at dinner. Since he suffers from memory impairment he

cannotalwaysrememberthat.Early inthemorning,before leavinghome,Alberto’sdaughter

subdivides theproperamountof thedifferentdrugs thathave tobe takenduring theday in

threeboxeswithdifferentcolour.At10:00a.m.thetouchscreenintegratedonthewheelchair

shows a reminder and sound alarm inviting Alberto to take his pills. The corresponding

intelligenttutorialshowstheinstructionsonhowtotakehismedication.Eventually,thesystem

asksAlbertotoconfirmtohavetakenhisdrugs.

Stuctureofthescenario:

Disability Degree Disease ScenarioFunctionalCategory

ActivityofDailyLiving

Cognitive Mild Alzheimer Home Medical Takemedication

Role of the IT service: thepatient‐agent detects that therapy timehas arrived and shows a

reminderfollowedbythecorrespondingtutorialonhowtotaketheprescribedmedicationon

the touch screen. Once the the tutorial workflow is completed, a confirmation message is

shownaskingtheusertoconfirmthatthemedicationhasbeentaken.

RoleofCommunications:Albertorequiresawheelchairtomovearound,equippedatleastwith

on‐boardPCconnectedtotheenvironmentviaWi‐Fi.


8.1.2Scenario2:Pietrogoestothechurch

Purposeofthescenario:typicalinteractionofauserwithamilddegreeofmixedimpairment

movingfromanindoortoanoutdoorenvironmentonthewheelchair.

UserDescription:Pietroisan80yearsoldmanwholiveswithaninformalcaregiver.Hesuffers

fromamildmixeddisabilityafterstrokewithahemiparesisand impairment intheexecutive

function.

NarrativeScenario:Twoyearsago,Pietrosufferedfromastroke.Herecoveredwellandatthe

momentheneedsonlya lightsupporttowalk. Sometimeshe isnotconfidentenough inthe

sequenceofactionsneededtoreachagoal.TodayisSundayandhiscaregiverhashisdayoff.

AteleveninthemorningPietrogetsareminderthatitistimetogetreadytogotothechurch.

Tutorials on how to put on his shoes and overcoat are displayed on his wheelchair touch

screen.




Mixed MildPost‐strokeHemiparesis

Home DressingPutonshoes

Putonovercoat

RoleofITservice:thepatient‐agentengagingalltheservicesavailablefortheuserrequiresthe

home‐agentthesuitedtutorialsforthesituation.Areminderislaunchedpriortotheexecution

ofthetutorialaboutsomebasicissuesrelatedtoprerequisite:

1. Areyouwearingyourshoes?

2. Areyouwearingyourovercoat?

RoleofCommunications:alldevicescommunicatewiththehome‐agentandtheITviaWi‐Fi.


8.1.3Scenario3:Francescacookspasta

Purpose of the scenario: typical interaction between the system and a user affected by a

moderatedegreeofmixedimpairmentonadailyroutine.

UserDescription:Francesca isa67yearsoldfemalewho livesathome,assistedbyChiara,a

professionalcaregiver.Shehasamoderatemixeddisabilitypost‐stroke.Lastyearshesuffered

fromastrokeofthelefthemisphere:shecannotwalkbutshe’sabletouseherarmsandshe

usesawheelchair.Shesuffersfromaphasiaaswell.

NarrativeScenario:TodayisSundayandChiaraisonherdayoffuntillateafternoon.Francesca

andChiaradecided,yesterday,tohavemaccheronialsugoforlunch.Beforeleavingthehouse,

Chiara arranged all the ingredients and kitchenware (pots, dishes, etc) to cook pasta on the

kitchentable.At12a.m.Francesca ishungry.Shemovesfromthe livingroomtothekitchen

driving her wheelchair on autonomous navigation mode. Francesca ask for help in meal

preparationtothesystemthroughthetouchscreen.Thecorrespondingtutorialshowshowto

cookpastastepbystep.Eachgraphicalrepresentationexactlycorrespondstotherealthingson

thetable,otherwiseaphasiawouldpreventherfromrecognizingtheexactobjects.




Mixed ModeratePost‐strokeAphasia

Kitchen Cooking Preparingpasta

RoleoftheITservice:oncetheuserreachesthekitchen,thepatient‐agent,afterengagingthe

ITservice,requeststhehome‐agenttheappropriatetutorialforthesituationthatwillofferthe

userhelponhowtopreparethemealwiththe ingredients.The information isshownonthe

suitableformataccordingtotheuserprofilestoredbytheagent.

Role of Communications: since Francesca drives to the wheelchair on her own, the

environment‐agent checks if she is in the kitchen viaWi‐Fi and thepatient‐agent offers the

tutorial to cook pasta on request. The touch screen integrated with the wheelchair


communicates directly with its local PC.Wi‐Fi can also be used to download cooking pasta

instructionsintothelocalPC.

Thefollowingtwoscenarioshavebeenspecificallycreatedforthe ITspecification inorderto

includetutorialsfromthosecategoriesnotyetconsidered.Bothscenarioshavebeenreviewed

bythemedicalteam.

8.1.4Scenario4:Annadoesthehousework

Purposeofthescenario:thisscenarioshowsthetypicalinteractionofauserwithamilddegree

ofcognitiveimpairment,duetotheevolutionofherdisease.

UserDescription:Anna isa75yearsoldwho liveswithher80yearsoldhusband.Hesuffers

fromdiabetesandarthritisbutisinagoodcondition.Annafelldownsixmonthsagowhenshe

wasleavinghomeandneedsawheelchairtomovearound.Shehadanoperationtwomonths

agoandishavingrehabilitationtwiceaweek.Sheissufferingfromdementiaonaearlystage

andisunabletorecognizesoundsduetoagerelateddeafness.

NarrativeScenario:Annahaddinnerwithherhusbandandshewantstowashthedirtydishes.

Sheissufferingfromtheearliestsymptomsofherdementiasosheforgotshowtogoonwith

tasks at certain point. She can barely hear anything with the right ear. His husband has

gatheredallthewashingitemsonthekitchencounternexttothesinkandhasputallthedirty

dishesononeside.TheintelligenttutorialshowsAnnahowtowashthedishesonrequest.




Mixed MildDementiaDeafness

HipfractureKitchen Housekeeping Washingdishes


Role of the IT service: thepatient‐agent detects dinner time has passed and Anna is in the

kitchen,sooffersherthewashingdishestutorialamongothers.Asitknowstheuserfunctional

profile,itautomaticallyofferstheconfigurationoftheassistantwiththevoicemodeoff.

Roleof Communications: this scenario involve that both theenvironmentandpatient agent

andthewheelchairareconnectedviaWi‐Fi,sothattheITservicegetsaninputofthecurrent

placeandtime.Annainteractswiththewheelchairviaatouchscreen.

8.1.5Scenario5:Stefanoreceivesthevisitofafriend

Purposeofthescenario:thisscenarioshowsthetypicalinteractionofauserwithamoderate

degreeofcognitiveimpairmentwiththeITservice.

User Description: Stefano is a 73 years old suffering from Parkinson at a primary stage. His

tremorand rigidityare becomingevident.Difficultieswithabstract thoughtandmemoryare

affectinghimtoo.HeliveswithaninformalcaregiverwhohastakenSaturdaymorningoff.

Narrative Scenario: In an hour, Stefano will receive the visit of a close friend at home. His

caregiver has helped him getting out of bed and has prepared himbreakfast before leaving.

Next,heneedstotidyhimselfup.Hehasgoodcoordinationbuthasmotorproblemsbecause

ofhisstiffness.Histremor islittleperceptiblebutitisnotaphysicalhandicap.Whenhegoes

intothetoiletwiththewalker,apackoftutorialstobeperformedonthatcontextshowsup:

washing hands and face, brushing teeth, shaving and hair combing. The intelligent tutorial

guides him step‐by‐step on how to complete the tasks considering hismotive condition and

tendencytodisorientation.




Cognitive Moderate Parkinson Bathroom Grooming

Washingface,handsBrushingteeth

ShavingHaircombing


RoleoftheITservice:thepatient‐agentdetectsStefanoisinthebathroomatearlymorning,so

offershimthepackofgroomingtutorialsamongothers.Astheserviceknowsabouthismotive

condition, instructions are displayed to facilitate the performance of the activity as long as

Stefanostickstotheflowofactions.

RoleofCommunications:thisscenarioinvolvetheenvironmentagentandi‐Walkerconnected

viaWi‐Fi,sothatthe ITservicegetsan inputofwheretheuser is.Stefano interactswiththe

tutorialviaatouchscreen.

8.2Activities’Description

EachofthescenariosdescribedintheprevioussectionincludesseveralADLstobeperformed.

It is necessary to remark that users suffering from a severe degree of disability cannot use

tutorialsduetotheirphysicalandmentalcondition.

Next,activitiesinwhichtutorialsarebasedaredetailedbythefollowingpoints:

Activityname: identificationstringfortheactivity.

Activitydefinition: whattheactivityisabout.

Activation: Actionsthattriggerthestartofthetutorial.

Itemsneeded: materialneededfortheproperperformanceoftheactivity.

Preconditions: specificationofwhatneedstobearrangedbeforethetaskstarts.

Activityworkflow: specificationofthestep‐by‐stepactionscomposingtheactivity.

Extensions: Alternativestothedefinedworkflowfortheactivity.

Contraindications: descriptionofthoseuserswhoarenotallowedtocarryoutthisactivityforsafetyreasons.


8.2.1Domestics

Domestics include those housekeeping activities demanding low effort from users. Activities

suchassweepingorironinghavebeenruledoutduetotherequiredphysicalcapabilities.

Activityname: WASHINGTHEDISHES

Activitydefinition:

Itconsistsofsoapingandrinsingthedirtydishes.Thecleandishesareputonthedrainingboardtogetdry.

Activation:Thistutorialisofferedafterdailymeals:breakfast,lunchanddinner.

Itemsneeded: Latexgloves,scourer,soap,plug,dishcloth,plasticdrainingboard.

Preconditions:• Alltheitemsmustbearrangedatthekitchencounter

• Itissupposedthesinkhastwosides

• Thedrainingboardmustbeempty

Activityworkflow:

1. Makesurenoneoftheseitemsismissing:latexgloves,ascourer,soap,aplug,adishcloth,aplasticdrainingboard

2. Putonthelatexgloves

3. Placeallthedirtydishesattherightsideofthesink

4. Blockthesinkwiththeplug

5. Turnonthetaptofilltheleftsideofthesinkwithwater

6. Turnoffthetapwhenthesinkishalffilled

7. Pouratrickleofsoapintothewatersoitgetssoapy

8. Placeallthedirtydishesattheleftsideofthesink,intothewater

9. Pouratrickleofsoaponthescourer

10. Softlyrubeachdishwiththesoapyscourerandplaceitontherightsideofthesink.Repeatstep10untilthereisnodishleft.

11. Turnonthetapandrinseeachdishovertheleftsideofthesink,placingitcarefullyontheplasticdrainingboard.Rinsefirstcupsandglasses,thenplatesandcutlery,andlastlypans,potsandsimilarkitchenware.Repeatstep11untilthereisnodishleft.

12. Drythesinkborderswiththedishcloth

13. Takeoffthelatexgloves


Extensions:

Oneormoreitemsaremissing(1)

1. Reportanalert

2. EndofUseCase

Allthecleandishesdonotfitonthedrainingboard(11)

1. Carefullyplacetheremainingdishesontherightsideofthesink.

Oneofthedishesbreaksintopieces(10,11)

1. Reportanalert

2. EndofUseCase

Contraindications:

Itisnotindicateforthoseuserssufferingfrom:

• ipovisionforsafetyreasons(e.g.notseeingclearlyknives)• essentialtremors

• usersaffectedbyhemiparesisthatspeciallyaffectstrenghandsensibilityonhandsandfingers

Activityname: WASHINGCLOTHESBYHAND

Activitydefinition:

Itconsistofwashingyoursoiledclothesbyhandandhangthemuptodry

Activation: Thistutorialisjustoffereduponrequest.

Itemsneeded: abasin,soiledclothes,liquiddetergent

Preconditions:

• Alltheitemsmustbearrangedatthekitchencounter

• Itissupposedthereisaplacetohandtheclothingitemstodry

• Iftheusersuffersfromsensitivitydeficit,itisrequiredtousecoldwater

Activityworkflow:

1. Makesurenoneoftheseitemsismissing:abasin,soiledclothes,liquiddetergent

2. Addallthesoiledclothestothebasin

3. Fillthebasinwithwarmwater

4. Pourliquiddetergentrightonthebasin

Foreachclothingitemyouneedtowash,repeatstep5:

5. Takeabumpysurfaceandrubitupagainsttheitem


6. Lettheclothessoakforabout20‐30minutes

7. Putyourhandsinthewaterandmooshtheclothingaround

8. Onceyouaredonemoooshing,drainoutthewater

Foreachclothingitem,repeatstep9‐10:

9. Rinseeachiteamundertherunningwatertillallthesoapisout

10. Handituptodry

Extensions:


1. Reportanalert

2. EndofUseCase

Contraindications: Thereisnospecificcontraindications

8.2.2Dressing

Thiscategoryisspecificfortutorialswhichhelpuserstogetdressedinordertogoout.Thereis

aspecialtutorialonhowtoputshoesonaimedatusersrequiringspecialneeds.

Activityname: GOINGOUTDOORS

Activitydefinition:Itconsistsofgettingreadytogoout.Theuserputsonhisshoes,putsonajacket/overcoatandfinallylocksthedoor.

Activation:Thetutorialisofferedwhenevertheuserisgoingoutandthesystemhasthisfactregistred.

Itemsneeded: Chair,shoes,socks(ifnecessary),cardiganorjacket,homekey.

Preconditions:

• Alltheitemshavetobearrangedontheuser’sbedroom.Eitherthecardiganorthejacketaresupposedtobehangingontheclothinghookoronavisibleplacewhereitwouldbeeasytoget

• Inordertoputontheshoescorrectly,onlyflatshoesandtrainerswithoutlacesareconsideredtobevalid

• Theuser’swheelchairwouldbeacceptedinsteadofthechair

• Theuserisnotsupposedtohaveanorthopediclegorarm

• Thecardigan/jacketdoesnotneedtobedoneup


Activityworkflow:

1. Makesurenoneoftheseitemsismissing:achair,socks(ifnecessary),shoes,acardiganorjacket,maindoor’skey

2. Sitdownonthechairandputontheshoes

3. Standupandputonthecardiganorjacket

4. Takethemaindoor’skey

5. Whenleavinghome,lockthemaindoorusingthekey

6. Keepthekeyonyourpocket

Extensions:


1. Reportanalert

2. EndofUseCase

Theuseriswearingslippers(2)

1. Removeyourslippersandplacethemaside

Theuserneedstouse/changesocks(2)

1. Takeoffbothyourleftandrightwornsocksifwearingany.Placethemaside

2. Putonthecleansocks

Theuserisnotalreadywearingashirt/t‐shirt/cardigan(3)

1. Reportanalert

2. EndoftheUseCase

Contraindications: Tutorialnotaimedatuserswhocannotgooutdoors.

Activityname: PUTTINGSHOESON

Activitydefinition:Itconsistsofsittingdownandputtingontheshoeswiththehelpofaspecialsplintandashoeborn.Thistutorialisaimedatusersrequiringoccupationaltherapy.

Activation:Thetutorialisofferedwhenevertheuserisgoingoutandthesystemhasthisfactregistred.

Itemsneeded: Chair,shoes,splint,socks,shoeborn.


Preconditions:

• Alltheitemshavetobearrangedontheuser’sbedroom

• Inordertoperformthetaskcorrectly,onlyflatshoesandtrainerswithoutlacesareconsideredtobevalid

• Theuser’swheelchairwouldbeacceptedinsteadofthechair

Activityworkflow:

1. Makesurenoneoftheseitemsismissing:chair,shoes,splint,socks,shoeborn.

2. Sitdownonthechair

3. Coverallthesplintwiththeleftsock

4. Putthetipofyourleftfootinsidethesockandslipitinsideusingthesplint’sstrips

5. Taketheshoebornanduseittohelpyouslipyourleftfootintotheshoe

6. Coverallthesplintwiththerightsock

7. Putthetipofyourrightfootinsidethesockandslipitinsideusingthesplint’sstrips

8. Taketheshoebornanduseittohelpyouslipyourrightfootintotheshoe

Extensions:


1.Reportanalert

2.EndofUseCase

Theuseriswearingslippers(2)

1. Removeyourslippersandplacethemaside

Contraindications: Thereisnospecificcontraindications.

8.2.3Cooking

Different levels of difficulty in terms of coordinationare required depending on the cooking

activity,sotherangeoftutorialswithinthiscategoryvariesaccordingtotheuserprofile.Basic

cookingskillsareinvolvedasallthetasksdetailedareeasytoperform.


Activityname: PREPARINGCOFFEE

Activitydefinition: itconsistsofpreparingcoffeeforanespresso.

Activation:thetutorialisofferedwhenevertheuserisonthekitchenanditisamorningoreveninghour.

Itemsneeded:Anespressomaker,ajugfilledwithwater,4spoonsofgroundcoffee,acoffeecup,acoffeespoon,analarmclock,sugar,acartonofmilk.

Preconditions:


• Theespressomakermustbedetachedinto3pieces

• Itissupposedthatthecookeriselectric

• Theuserknowshowtoturnonthecookerandsetthealarm

Activityworkflow:

1. Makesurenoneoftheseitemsismissing:anspressomaker,ajugfilledwithwater,groundcoffee,acoffeecup,acoffeespoon,analarmclock,sugar(ifnecessary),acartonofmilk(ifnecessary)

2. Fill¾oftheespressomakerbasewithwater

3. Fitthepiecewithlittleholesontopofthebaseandfillitwiththegroundcoffee

4. Screwthetworemainingpiecestightly

5. Turnonthefurthestceramichoboftheelectriccookerandplacethespressomaker

6. Setthealarmclocktoringafter5minutes

7. Checkifthetopoftheespressomakerisfilledwithcoffee.Repeatthissteptillallthecoffeehascomeuptothetoporthealarmclockrings.

8. Whenthecoffeehascomeuptothetop,turnofftheceramichob

9. Carefullypourthecoffeeonthecup

Extensions:


1. Reportanalert

2. EndofUseCase

Theuserlikesasweetercoffee(9)

1. Addthe2teaspoonsofsugartothecoffee

Theuserlikessomemilkinthecoffee(9)

1. Addatrickleofmilktothecoffee


Contraindications:


• ipovisionforsafetyreasons(e.g.notseeingclearlyknives)

• essentialtremors


• userssufferingfromdysphagia

Medicalassesmentshouldindicatethelevelofdeficitlimitingthisactivity.

Activityname: COOKINGPASTA

Activitydefinition: itconsistsofboilingfreshpastaandprepareittobeserved.

Activation:thetutorialisofferedwhenevertheuserisonthekitchenanditistimetoprepareameal.

Itemsneeded:100gramsoffreshpasta,salt,sauce,1literofwater,apot,scissors,acolander,awoodenspoon,afork,analarmclock,kitchenmitten,aplate.

Preconditions:



• Itissupposedtheuserknowshowtoturnonthecookerandsetthealarmclock

Activityworkflow:

1. Makesurenoneoftheseitemsismissing:100gramsoffreshpasta,salt,sauce(ifnecessary),1literofwater,apot,scissors,acolander,awoodenspoon,afork,analarmclock,kitchenmitten,aplate

2. Fillthepotwith1literofwater

3. Turnonthefurthestceramichoboftheelectriccookerandplacethepottoheatup

4. Addadessertspoonofsalttothewater

5. Openthepacketofpastawithascissors

6. Whenthewatercomestotheboil,addthe100gramsoffreshpasta

7. Stirthewaterwiththewoodenspoonfromtimetotime

8. Setthealarmclocktoring5minutesafterthewaterboilsagain

9. Placethecolanderononesideofthesinkandopenthetapalittlebit


Activityworkflow:

1. Makesurenoneoftheseitemsismissing:100gramsoffreshpasta,salt,sauce(ifnecessary),1literofwater,apot,scissors,acolander,awoodenspoon,afork,analarmclock,kitchenmitten,aplate

2. Fillthepotwith1literofwater


4. Addadessertspoonofsalttothewater

5. Openthepacketofpastawithascissors

6. Whenthewatercomestotheboil,addthe100gramsoffreshpasta

7. Stirthewaterwiththewoodenspoonfromtimetotime

8. Setthealarmclocktoring5minutesafterthewaterboilsagain

9. Placethecolanderononesideofthesinkandopenthetapalittlebit

10. Whenthealarmclockrings,turnofftheceramichob

11. Carefullypourthecontentofthepotintothecolanderusingthekitchenmitten

12. Whenthepastaisdrained,spreaditontheplate

Extensions:


1. Reportanalert

2. EndofUseCase

Theuserlikestoaddsaucetothepasta(12)

1. Addasmuchsauceasyoulikeusingthespoon

Contraindications:







Activityname: MAKINGASANDWICH

Activitydefinition: itconsistsofpreparingasandwichmadeofhamandmozarella.


Itemsneeded:4slicesofham,4slicesoffreshbuffalo‐milkmozzarella,2slicesciabatta(1inchthick),aplate,basil.

Preconditions: Alltheitemsmustbearrangedatthekitchencounter

Activityworkflow:

1. Makesurenoneoftheseitemsismissing:4slicesofham,4slicesoffreshbuffalo‐milkmozzarella,2slicesciabatta(1inchthick),aplate,basil

2. Placethehamononeciabattasliceandthemozzarellaontheother

3. Garnishwithbasil

4. Closethesandwichandplaceitonaplate


Extensions:


1. Reportanalert

2. EndofUseCase

Contraindications: Itisnotindicateforthoseuserssufferingfromdysphagia

Activityname: PREPARINGCOLDSOUP

Activitydefinition: itconsistsofpreparingacoldvegetablesoup.


Itemsneeded:1literofchickenstock,2bayleaves,groundpepper,apot,30gramsofnoodles,awoodenspoon,analarmclock,aladle,asoupdish.

Preconditions:



• Itissupposedtheuserknowshowtousetheelectriccookerandsetthealarmclock

Activityworkflow:

1. Makesurenoneoftheseitemsismissing:1literofchickenstock,2bayleaves,groundpepper,apot,30gramsofnoodles,awoodenspoon,analarmclock,aladle,asoupdish

2. Fillthepotwith1literofchickenstock,2bayleavesandthegroundpepper


4. Whenthechickenstockcomestotheboil,addthe30gramsofnoodles

5. Setthealarmclocktoring3minutesafterthesoupboilsagain

6. Stirthesoupwiththewoodenspoonfromtimetotime

7. Whenthealarmclockrings,turnofftheceramichob

8. Servethesoupintothedishusingtheladle

Extensions:


1. Reportanalert

2. EndofUse


8.2.4Medical

Medical tutorial is aimed at those userswho need to take prescibed drugs daily to alleviate

theirdiseases’symptoms.Inordertoseparatethedosesthatneedtobetakenalongtheday,

differentcolouredboxesareused(e.ggreenboxformorningdoseandyellowboxforevening

dose).

Contraindications:







Activityname: TAKINGDRUGS

Activitydefinition:itconsistsofindicatingusersitistimetotaketheprescribeddoseandguidethemonhowtodoit.

Activation:thetutorialisofferedwhenevertheuserrequirestotakeadoseduetomedicalprescription(generallyinthemorning,aftermiddayorintheevening).Itsactivationisprecededbyareminder.

Itemsneeded: specificlabeledbox,aglassofwater.

Preconditions: caregiversareinchargeoffillingthelabeledboxeswiththerightdoseaccordingtotheuser’sprofile.

Activityworkflow:

1. Makesurenoneoftheseitemsismissing:specificlabeledbox,aglass,abottleofwater

2. Grabthered*labeledboxandtakeoutthetop

Foreachofthepillsyouhavetotake,repeatsteps3‐5:

3. Takeapillandputitintoyourmouthbutdonotswallowityet

4. Takeaglassofwaterandhaveadrinktoswallowthepill.Haveanotherdrinkofwatertomaketheingestioneasier


8.2.5Grooming

Groomingtutorialsareessentialtoenhancetheself‐esteemandindependenceofusers,since

the tasks to perform have a direct effect on themselves. The morning grooming tutorial is

offeredaftertheuserwakesup.

Activityworkflow:

1. Makesurenoneoftheseitemsismissing:specificlabeledbox,aglass,abottleofwater

2. Grabthered*labeledboxandtakeoutthetop

Foreachofthepillsyouhavetotake,repeatsteps3‐5:

3. Takeapillandputitintoyourmouthbutdonotswallowityet

4. Takeaglassofwaterandhaveadrinktoswallowthepill.Haveanotherdrinkofwatertomaketheingestioneasier

5. Putthetoponthecolouredbox

5. *Coloursaremeanttodistinguishdifferentdosesthroughtheday

Extensions:


1. Reportanalert

2. EndofUseCase

Contraindications:Daltonicsarenotabletousethetutorial,sincecoloursidentifythepillboxes.

Activityname: MORNINGGROOMING

Activitydefinition:

itconsistsofasetofactivitiestoperformintheearlymorning:washinghandsandface,brushingteeth,haircombingandshaving(ifnecessary).

Activation:thetutorialisofferedwhenevertheuserentersthebathroomforthefirsttimeonaday.

Itemsneeded:pumpdispenser,cleanser,atowel,amanualtoothbrush,toothpaste,aplasticcup,ahaircomb,anelectricshaver

Preconditions:

• Alltheitemsmustbearrangedclosetothewashbasin

• Iftheusersuffersfromsensitivitydeficit,itisrequiredtousecoldwater

Activityworkflow:

1. Makesurenoneoftheseitemsismissing:apumpdispenser,acleanser,atowel,amanualtoothbrush,toothpaste,aplasticcup,ahaircomb,anelectricshaver.

2. Getsoapfromthepumpdispenserandwashyourhands

3. Washyourfaceusingthecleanser


Activityworkflow:

1. Makesurenoneoftheseitemsismissing:apumpdispenser,acleanser,atowel,amanualtoothbrush,toothpaste,aplasticcup,ahaircomb,anelectricshaver.

2. Getsoapfromthepumpdispenserandwashyourhands

3. Washyourfaceusingthecleanser

4. Wetyourhandsusingthetowel

5. Squeezesometoothpasteontothetoothbrushandbrushyourteeth

6. Filltheplasticcupwithwaterandhaveasiptorinseyourmouth

7. Combyourhairusingthehaircomb

8. Usetheelectricshavertoshaveyourstubble

Extensions:


1. Reportanalert

2. EndofUseCase

Contraindications:


• ipovisionforsafetyreasons(e.g.notseeingclearlytherazor)




Activityname: BRUSHINGTEETH

Activitydefinition:

Itconsistsofcleaningoneselfteethusingtoothpaste.Thistutorialisaimedatusersrequiringoccupationaltherapy.

Activation: thetutorialisoffereduponrequest.

Itemsneeded: amanualtoothbrush,toothpaste,aplasticcup.

Preconditions:

• Alltheitemsmustbearrangedclosetothewashbasin.

• Iftheusersuffersfromsensitivitydeficit,itisrequiredtousecoldwater.

Activityworkflow:

1. Makesurenoneoftheseitemsismissing:atoothbrush,toothpaste,aplasticcup

2. Squeezesometoothpasteontothetoothbrush

3. Turnonthetaptowetthetoothpastealittle

4. Useshort,back‐and‐forthbrushingmotionstocleantheoutsideandinsidesurfacesoftheteethwiththetoothbrush,aswellasthechewingsurfaces


Activityworkflow:

1. Makesurenoneoftheseitemsismissing:atoothbrush,toothpaste,aplasticcup

2. Squeezesometoothpasteontothetoothbrush

3. Turnonthetaptowetthetoothpastealittle

4. Useshort,back‐and‐forthbrushingmotionstocleantheoutsideandinsidesurfacesoftheteethwiththetoothbrush,aswellasthechewingsurfaces

5. Brushalongthegumline.Makesuretobrushyourbackmolarsandtonguetoremovebacteria

6. Turnonthetaptofilltheplasticcupwithwater

7. Haveasipofwatertorinseyourmouth

8. Spitoutthewaterintothewashbasin.Repeatstep5to7again.

9. Turnonthetaptocleanupthewashbasinandyourtoothbrush

Extensions:


1.Reportanalert

2.EndofUseCase

Contraindications:






9.REQUIREMENTS

9.1Actors’interaction

ITareaimedatusersbeingabletoperformdailybasicactivitieswiththecaregivers’checking.

Themedicaluserprofile,definedbydoctorsandpediatricians,isupdatedeverytimetheusers’

capabilitieschange.TheserviceispartofthePatientAgent,wherethecontextualinformation–

locationoftheuseratanymoment‐isgatheredfromtheenviromentsensors.

The following diagram shows all the possible interaction flows among entities (actors and IT

service)thatcanoccurwithinthesystem.

Figure9.InteractionflowsamongactorsandtheITservice


9.2FunctionalRequirements

RequirementsareallthoseneedsthatkeyactorshaveinthecontextwhereITwillbeused.All

theserequirementstogetherdescribethebehavioroftheserviceandthefunctionalitiesthat

thesystemissupposedtoaccomplish.

Requirement: SELECTATUTORIALUPONREQUEST

Description:Thesystemshallallowtheusertoselectanyofthetutorialsavailable

Rationale: TheaimoftheserviceistofacilitateusersperformanceofcertainADLs

Fitcriterion: Theuserisabletoselectanytutorialaddressedtohim/her

Requirement: TUTORIALCATEGORIZATION

Description: Tutorialsaregroupedintocategoriesaccordingtotheirrelation

Rationale: Tutorialsareclassifiedinordertosimplifytheselectionofactivitiestocarryout

Fitcriterion:Theusernoticethereisseveralsubsetsoftutorialsdependingontheircommonsubject

Requirement: STEPSCONFIRMATION

Description:Everytutorial’sinstructionisvalidatedaskingtheuserwhethers/hehasunderstoodtheindicationsgiven

Rationale: Comprehensionofeachstepcomposingthetutorialisguaranteedbyusingconformationmessages

Fitcriterion:Theuseracknowledgess/hehassuccessfullyperformedtheinstructiongivenandaccepttheconfirmationmessage

Requirement: TUTORIALBUTTON


9.3Non‐FunctionalRequirements

Non‐functional requirements are those describing criteria that can be used to judge proper

operationofIT,ratherthanspecificbehaviours.Non‐functionalrequirementsdefineconstraints

onthedesignandimplementationofITservice.

Description: TheGUImustincludea“tutorial”button

Rationale: The“tutorial”buttonallowstheusertodisplaythelistofcategorizedtutorialsthats/hecanperform

Fitcriterion:Theuseracknowledgesthereisavisibleandrecognizablebuttontoaccesstutorials

Requirement: HELPBUTTON

Description: TheGUImustincludea“help”button

Rationale: The“help”buttonallowstheusertotransfertheexplicithelpquerytotheHelpAssistantManagementservice

Fitcriterion: Thecaregiveracknowledgesthattheuserisaskingforextrahelp

Requirement: QUITBUTTON

Description:TheGUImustincludea“quit”buttonwhendisplayinganytutorialstep

Rationale: The“quit”buttonallowstheusertoexitthecurrenttutorialatanystep

Fitcriterion:Theuseracknowledgesthereisavisibleandrecognizablebuttontoquitatutorialwhilebeingactive.Thesystemdisplaysaconfirmationmessagewhenthebuttonisclicked.


9.3.1Look&Feelrequirements

9.3.2Usabilityrequeriments

9.3.3Operationalrequirements

Requirement: SIMPLEANDEASYSERVICE

Description:Theinterface,whichisinchargeofdisplayingtheITservice,mustbeuserfriendlytoallkindofusersconsideredtobeinteracting.

Rationale:ItisessentialthattheinterfaceiseasytousesinceITarebasedontheinteractionwithusershavingdisabilities.Ithastobeadaptabletotheuser’scognitivecondition.

Fitcriterion:Anyusershallbeabletoknowhowtointeractwiththesystemandhowtorequestservicesinlessthanthreetries.

Requirement: USABLESERVICE

Description:Thesimplicityofthetutorialsshallmakeuserstohaveagoodperceptionofthesysteminordertobewillingtouseitinfutureoccasions.

Rationale:Bynorm,userspreferserviceseasytouseandwhichdonotimplyanunnecessarywastetime.

Fitcriterion:Finalusersfindtutorialseasytouse.Theydonotneedtohavepriortechnologicalexperience.

Requirement: FASTDISPLAYOFDATA

Description: Theresultofaqueryshallbeshowedasfasteraspossible.

Rationale: Finalusermustobtaindatainafastway.

Fitcriterion: Thesearchingresponsetimewillbeinferiorto0.5seconds.

Requirement: ACCESSIBILITY


9.3.4SupportandMaintenancerequirements

Description: Thesystemshallbeaccessible24haday,everydayoftheyear.

Rationale:ThesystemmustbeavailableallthedaytobeabletoperformanyoftheADLsusingITatanymoment.

Fitcriterion: Thesystemwillbealwaysavailable.

Requirement: SET‐UPREQUIREMENTS

Description: Noexternalset‐uprequired.

Rationale: Thesystemmustbeeasytointegrate.

Fitcriterion: Theuserwillbeabletousethesystemwithoutinstallinganyextrasoftware.

Requirement: SCALABLESERVICE

Description: Thesystemshallbescalableorextendedeasily.

Rationale:Thesystemshouldbeabletodealwithaconsiderablenumberoftutorials.

Fitcriterion: Thesystemrunsproperlywhenbeingextended.

Requirement: DATAINTEGRITY

Description:Thesystemmustguaranteethedataintegrityofanyinformationintroducedbyusers.

Rationale:Thesystemmustavoiddatamistakesandanyinconsistencywhichmayaffectusers.

Fitcriterion: Thesystemmustbeconsistentwithdatacontentatanytime.


9.4UseCases

9.4.1UseCasesDiagrams

UseCasesdescribehowactors interactwiththe systeminordertoachieveagoal,specifying

whocanexecuteacertainactiontowardstheITservice.Followingdiagramsshow,foreachof

the users contemplated, the possible actions to carry out in order to fulfill functional

requeriments.

Figure10.Actor’sspecializationdiagram

Figure11.UseCasesrelatedtotheUseractor


Figure12.UseCasesrelatedtotheSystem

Figure13.UseCasesrelatedtotheAdministratoractor

SinceDoctors/Geriatriciansupdatetheuserprofileusinganotherservice,thesubsequentUse

Case isnot relevant for thespecification.Caregiversdonot interactdirectlywith thesystem,

they are responsible for assuring tutorials’ preconditions and instruct users in how the

applicationworks.


9.4.2UseCasesSpecification

Following,thecompletespecificationofeachUseCasefromthediagramsaboveispresented,

detailingpreconditionstobefulfilledbeforeexecutingtheactionandhowtovalidatethatthe

goalhasbeenachievedsuccessfully.

UseCase: REQUESTTUTORIAL

Description: Theuserrequestacertaintutorialtobeperformed

PrimaryActor: User

Preconditions:Preconditionsdefinedfortheactivityoftheresquestedtutorialmustbefulfilled

Validationcriteria: Therequestedtutorialisinitiatedsuccessfullybytheuser

MainSuccessScenario:

1. TheuserclicksontheGUI’s“tutorials”button

2. Theuserchoosesthecategoryofthetutorial

3. Theuserchooseswhichtutorials/heisabouttoperform

UseCase: PERFORMTUTORIAL

Description:Theuserperformstherequestedtutorialconfirmingeachofthesteps

PrimaryActor: User

Preconditions:Preconditionsdefinedfortheactivityoftheresquestedtutorialmustbefulfilled

Validationcriteria: Therequestedtutorialisperformedsuccessfullybytheuser


Foreachofthetutorial’ssteps:

1. Theuserfollowstheinstructionsgiven

2. Theuserconfirmss/hehasperformedthecurrentstepinordertomoveintothefollowing

UseCase: QUITTUTORIAL


Description: Theuserexitsthecurrenttutorialatanystep

PrimaryActor: User

Preconditions: Thecurrenttutorialhastobestarted

Validationcriteria: Therequestedtutorialisquittedsuccessfullybytheuser


1. TheuserclicksontheGUI’s“exit”button

2. Theuserconfirmss/hewantstoescapethetutorial

3. ThesystemleadstheusertothemainmenuoftheGUI

UseCase: ASKFORHELP/REPORTANALERT

Description:Theuserasksforhelp/reportanalertwhileperformingatutorial.ThisusecaseconcernsafunctionalityimplementedbyanotherservicewithinSHARE‐itwhichcollaborateswithITservice.

PrimaryActor: User

Preconditions: Thecurrenttutorialhastobestarted

Validationcriteria: Thehelprequesthasbeenprocessedandtheuserisinformedthatthecaregiverwillcontacthim/herforassistance


1. TheuserclicksontheGUI’s“help”buttonatanyofthetutorial’ssteps

2. ThesystemtransferthequerytothecaregiverthroughtheHelpAssistanceManagementservice

3. Thesysteminformstheuserthats/hewillbecontactedbythecaregiverforassistance

UseCase: LISTTUTORIALS

Description:Alistoftutorialstheusercanperformaccordingtohis/herprofileisofferedthroughtheGUI

PrimaryActor: System

Preconditions: Theuserprofilehasbeeninitializedorupdated

Validationcriteria: Asetoftutorialsadaptedtotheuserprofileislisted



1. Alistoftutorialsaccordingtotheuser’sphysical/cognitiveconditionissentfromthePatientAgenttotheInterfaceManager

2. ThesystemshowsthelistthroughtheGUIwhenclickingonthe“tutorials”button

UseCase: OFFERCONTEXTUALTUTORIALS

Description:Asubsetoftutorialsaredisplayedtotheuserdependingonthecontext

PrimaryActor: System

Preconditions: Locationhastobechanged.

Validationcriteria: Certaintutorialsaredisplayedwhenevertheusercontextchanges


1. Thesystemdetectsthattheuser’slocationhaschanged

2. Thesystempreselectasubsetoftutorialsfortheusertobeperformedaccordingtothenewcontext

3. Thesystemdisplaysthesubsetoftutorialsregardlessofnoexplicitdemand

UseCase: SYSTEMMAINTENANCE

Description: Maintanceandupdateofthesystem

PrimaryActor: Administrator

Preconditions: Notspecified

Validationcriteria: Thesystemisadjustedtotheuserneedsandprofilechanges.


1. Thesystemadministratoracknowledgesthereareupdatestobeexecuted.

2. Updatesarecarriedout.

3. Thesystemadministratorinformsdoctorsandcaregiversofthechanges.


10.RISKANALYSIS

10.1NewProblems

Thereareseveralissuesaffectingusersandcaregiversthatmightappearaftertheystartusing

IT.Regardingtheseissuespriortoitsconsequencesisakeyfactorfortheirpromptresolution.

10.1.1UsersandCaregiversProblems

Users.Using tutorials instead of being guided and supported by a caregiver allows users to

become more independent. However, it adds a new intermediary between the user and

caregivers/doctors.Asaresult, it canbesignificantlymorecomplicatedforuserswhodonot

managenewtechnologytoaccepttheintroductionofanewassistantaspartoftheirtherapy.

Another important issue is the user perception of the system. It is likely that users

misunderstandinstructionsorperforminthewrongwayregardlessofthesystemaccuracy.In

ordertominimizeitsimpactonthetutorialperformance,helpisavailablefortheuseratanyof

thetutorial’ssteps.

Caregivers. IT activation needs certain efforts to bemade from caregivers in order towork

successfully.Itisprobablethatinitially,thesystemresultsunattractiveforgrown‐upcaregivers.

Therefore, the tutorials’ efficiency must be proven to be higher than using the traditional

methodtoassistusers.

10.1.2Follow‐upProblems

Thetutorials’usageentailsanupdateoftheuserprofile.Delaysontheprofilemodificationcan

leadtoadeprecatedsystem,whereend‐usersmightbeusingaservicewhichdoesnotfittheir

needsandmisconceivestheirphysicalandmentalcondition.


10.2Risks

A risk in the context of software is any damage that can occur anytime and affect the

developmentcourse.Sincerisksareharmfulbynature,theonlywaytominimizeitseffectsisto

detectthepotentialrisksduringthisearlystageoftheproject,analyzeitsimpact,andplanhow

todealwiththeminorder.

10.2.1Personnelshortfalls

ITcanfailduetopersonnelinteraction.Itcannotbeasumedthatalltheusersinvolvedinthe

useoftutorialswouldachievetheexpectedexperience.Inordertomitigatethisrisk,aproper

controloverthewholesystemisneededtodetectanddealwithsituationsinwhichactorsdo

notinteractasexpected.

10.2.2Developingthewrongfunctions

Thesebothrisksdealwithdevelopingthewrongservice.Itcouldbethattheresultingsystem

doesnotsuittherequestedfunctionalitiesorthattheuserinterfaceisnotappealingtothem.

Inordertomitigatetheprobabilityoftheserisks, it isnecessarytovalidatethefulfillmentof

usecases,functionalandnon‐functionalrequirements.Regardingtheuserinterface,itishighly

recommendedtoprototypetheinterfaceinordertobetestedbythefinalusers,sothatthey

canprovidefeedbackasearlyaspossible.

10.2.3Continuousstreamofrequirementschanges

If the system requirements are continually changing, the schedulewill never be accurate. A

change in the current requirements impacts all the development’s phases and add time and

resources to the plan. In order to deal with this type of situations, iterations where new

requirementsarecollectedmustbeestablished.


10.2.4Tableofrisks

Risk Probability Impact

Personnelshortfalls 50% Verycritic

Developingthewrongfunctions 30% Verycritic

Continuousstreamofrequirementschanges 30% Critic


IMPLEMENTATION_______________________________________

11.TUTORIALS’MODULERELATIONS

ITserviceispartofthePatientAgent.ItspresentationlayercorrespondstotheGUI,whichwas

implementedasanotherpartoftheSHARE‐itproject.ITserviceandGUIcommunicatesthrough

DLAConnectionsusingdatapackages.Following,thereisabriefexplanationoftheseconcepts.

• GUI.TheGrahicalUserInterfaceplaysanimportantroleonthetutorials’display,asitis

the main front end for users and the way they interact with the service content,

accesingtoitsfunctionalities.

• DLA.Itisanarchitectureusedtodevelopcooperativesystemswithdifferentinteracting

modules distributed along several machines. This mechanism of communication is

based on a central element thatmanages the access and storage of the shared data

from theothermodules.When the IT servicewants to send some information to the

GUIorviceversa,aDLAinput/outputconnectionisopenedandadefinedDLApackage

withtherequesteddataissent.

Following, therearedefined thedifferent interaction schemasconcerning IT servicebetween

GUIandthePatientAgentlayer:

LISTOFTUTORIALS–CONTEXTUALTUTORIALS

PatientAgent InterfaceManager GUI

Caregiver/Doctorselectsuserprofile

Senduserprofile

(qualifiers+quantifiers)


SelectListofTutorials(name+numOfSteps+N[index+steptext+mediatype+media])accordingtouserprofile

Displays

SelectsubsetofContextualTutorials(name)accordingto

Location Displays

Sendsconfirmationofstep(name+index+

update)

SELECTATUTORIAL

PatientAgent InterfaceManager GUI

Userselectsatutorial

Sendsrequest(name+update)

Sendstutorial(name+numOfSteps+N[index+steptext+mediatype+media])

Displays

Sendsconfirmationofstep(name+index+

update)

Several different DLA3 packages are in charge of representing interactions between the IT

serviceandtheGUIthroughthePatientAgent[20].

3RefertosectionA1.2tofindoutmoreaboutthisarchitecture


11.1DLAPackage:TutorialList

• FromRolland/GUItoPatientAgentandviceversa

• FromiWalker/GUItoPatientAgentandviceversa

• Sentwhenevertheuserprofileisinitializedorupdated

When the user profile is initialized or updated through the GUI, the intelligent layer of the

systemselectsalistoftutorialsfortheusertoperform.Theselectionismadeaccordingtothe

typeofconditionsaffectingtheuser,andissenttotheGUIfordisplay.

struct Tutorial { char tutorial[128]; // name char activityCategory[64] // category of the tutorial’s activity int numOfSteps; // number of steps for the tutorial TutorialStep steplit[25] // where ‘n’ refers to the number of // steps a tutorial has. Set to // 25 as max number } struct TutorialStep { char steptext[1024]; // text describing one tutorial step char media[64]; // the name of the media file (if any) // to be shown } struct TutorialList { Tutorial tutorials[11]; // Where 'n' refers to the number of // available tutorials. Set to 11 // arbitrarily }

11.2DLAPackage:SelectedTutorial

• FromRolland/GUItoPatientAgent

• FromiWalker/GUItoPatientAgent

• Sentwhenevertheuserrequestsatutorial


struct SelectedTutorial { char tutorial[128] // name of the tutorial int update; // [1,..., 2147483647] number that changes // from one selection to another to allow // identification of new selection }

11.3DLAPackage:ConfirmationTutorialSteps

• FromRolland/GUItoPatientAgent

• FromiWalker/GUItoPatientAgent

• Sentwhenevertheuserconfirmsatutorialstep

struct ConfirmTutorialStep { char tutorial[128] // name of the tutorial int step; // [1,..., 2147483647] number that changes // from one selection to another to allow // identification of new selection }

11.4DLAPackage:ContextualTutorials

• FromPatientAgenttoRolland/GUI

• FromPatientAgenttoiWalker/GUI

• Sentwheneverthesystemdetectsauseriswithinadefinedlocation

struct ContextualTutorials { int numOfContextTuts; // number of available contextual tutorials // tutorials TutorialName tutorials[11] // name of the tutorial } struct TutorialName { char name[128]; }


12.JADEX

12.1IntroductiontoJADEX4

JADEXisanagent‐orientedreasoningengineforwritingrationalagentswithXMLandtheJAVA

programming language. Thereby, JADEX represents a conservative approach towards agent

orientation for several reasons. One main aspect is that no new programming language is

introduced. Instead, JADEXagentscanbeprogrammed inthestate‐of‐theartobject‐oriented

integrateddevelopmentenvironments(IDEs)suchasEclipse.

Agents represent active components with individual reasoning capabilities. This means that

agents can exhibit reactive behavior (responding to external events) as well as pro‐active

behavior(motivatedbytheagentsowngoals).

12.1.1TheBDIModelofJADEX

The most interesting and widespread agent architecture is the Belief‐Desire‐Intention (BDI)

architecture, introduced by Bratman as a philosophical model for describing rational agents

[25].Itconsistsoftheconceptsofbelief,desireandintentionasmentalattitudesthatgenerate

human action. Rao and Georgeff [26] transformed this model into a formal theory and an

executionmodelforsoftwareagents,basedonthenotionofbeliefs,goals,andplans.

JADEXfacilitatesusingtheBDImodelinthecontextmainstreamprogramming,byintroducing

beliefs, goals and plans as first class objects that can be created andmanipulated inside the

agent.InJADEX,agentshavebeliefs,whichcanbeanykindofJavaobjectandarestoredina

BeliefBase.Goalsrepresenttheconcretemotivations(e.g.statestobeachieved)thatinfluence

anagent'sbehavior.Toachieveitsgoalstheagentexecutesplans,whichareproceduralrecipes

coded in Java. The beliefs, goals and plans of the agent are defined by the programmer and

prescribethebehavioroftheagent.

4http://jadex.informatik.uni‐hamburg.de/bin/view/About/Overview


Reasoning inJADEX isaprocessconsistingoftwo interleavedcomponents.Ontheonehand,

theagent reacts to incomingmessages, internaleventsandgoalsby selectingandexecuting

plans.Ontheotherhand,theagentcontinuouslydeliberatesaboutitscurrentgoals,todecide

aboutaconsistentsubset,whichshouldbepursued[10].

12.1.2AgentSpecification

Thecompletedefinitionofanagent iscaptured inasocalledagentdefinitionfile(ADF).The

ADFisanXMLfile,whichcontainsallrelevantpropertiesofanagent(e.g.thebeliefs,goalsand

plans).

To develop applications with JADEX, the programmer has to create two types of files: XML

agentdefinitionfiles(ADF)andJAVAclassesfortheplanimplementations.TheADFcanbeseen

as a type specification for a class of instantiated agents. To start an agent, first the ADF is

loaded,andtheagentisinitializedwithbeliefs,goals,andplansasspecified.

Figure14.ComponentsofaJADEXAgent


12.2Tutorials’JADEXclasses

Inthenextsubsections,theBDImodelisexplainedforthebeliefs,goalsandplansdefiningthe

PatientAgentbehaviourforthetwofunctionalitiesoftheITservice:selectionofthetutoriallist

and contextual tutorials. Several different java classes are executed as plans (means‐end

reasoning)tohandleincomingeventsandtopursueinternalgoals.

12.2.1ListofTutorials

ThelistoftutorialsausercanperformissenttotheGUIthroughtheDLAconnectionwhenthe

user profile is either initialized or updated. This set of tutorials is selected according to the

physicalandcognitiveconditionstheuserpresents.

ProfileGUI.java.When the Patient Agent is launched, a JFrame for the caregiver/doctor to

introducetheuserprofileisdisplayed.Iftheacceptbuttonispressed,thecorrespondingfacts

ontheBeliefBaserelatedtotheuser’sconditions(visualimpairment,sound impairment,fine

coordination, aphasia, dementia, neglect) are updated. Consecutively, these new values are

transformedtobytesandsenttotheGUIthroughtheDLAConnectioninteractionProfileConn.

Following,anewgoal iscreatedanddispatchedinordertoselectthesuitable listof tutorials

for the user according to his/her profile. To reach this goal, a plan needs to be executed:

RuleEnginePlan.

RuleEnginePlan.java. Within this class, a set of rules composing the intelligent layer of the

systemarerunforthetutorials’selection.ThecorrespondingfactsontheBeliefBaserelatedto

the available tutorials (Grooming, PreparePastaSimple, PreparePastaComplex, PrepareCoffee,

GoingOut,Therapy)areupdatedoncetheselectionismade.

Following,anewgoaliscreatedanddispatchedinordertosendthelistofselectedtutorialsto

theGUIfordisplay.Toreachthisgoal,aplanneedstobeexecuted:TutorialListPlan.


TutorialListPlan.java. This class specifies in its body which are the actions to be taken for

initializingthelistoftutorialsfortheend‐user.Adatastructureisinitializedwhere,foreachof

the tutorials, the following attributes are contained: number of required steps, tutorial’s

category,thenameofthetutorial,anditssetofsteps.Althoughthesumoftutorialstheinitial

listcancontain issettomaximum11arbitrarily, itcanbe furtherextended.Whenthe listof

tutorials iscompleted, itsdatastructure is transformedtobytesandsenttotheGUIthrough

theDLAConnectiontutorialListConn.

Figure15.PatientAgent’sbehaviourrelatedtotheListofTutorials

12.2.2ContextualTutorials

Oneof themost important features regarding IT service is the possibility of offering certain

tutorials to the targetusersdependingonthecontext.The input informationofwhereusers

are located is retrieved from the network sensors via the Environment Agent, which closely

collaborateswiththePatientAgentforthispurpose.


WithintheBeliefBasetherearetwobeliefsrelatedtothecontextualization:thelocation,which

contains the name of the place itself taken from the ontology, and the location_change, a

booleanthatindicatesifthelocationhaschangedornot.

Twodifferenteventscantriggertheupdateofcontextualtutorials:

1.ChangeintheLocation.

SaveLocationPlan.java.Thisplanisinchargeofhandlingtheupdateofthelocationbeliefevery

timetheEnvironmentAgentinformstothePatientAgentthattheuser’slocationhaschanged.

Anewgoaliscreatedanddispatchedtoupdatethecontextualtutorials.Inordertoreachthis

goal,aplanneedstobeexecuted:UpdateContextualTutPlan.

Figure16.PatientAgent’sbehaviourwhentheuser’slocationchanges

2.Changeintheuserprofile.

TutorialListPlan.java. When the user profile is initialized or updated, the list of convenient

tutorials for the user to perform is selected. Consequently, as the contextual tutorials are a

subset of these tutorials, a new goal is created and dispatched in order to update the


contextual tutorials according to the current location of the user. To reach this goal, a plan

needstobeexecuted:UpdateContextualTutPlan.

Figure17.PatientAgent’sbehaviourwhentheUserProflilechanges

UpdateContextualTutPlan.java.Withinthisclass,thecurrentroomwheretheuserislocatedis

usedtodiscriminatethesubsetofcontextualtutorialstobeofferedtotheend‐user.Theinitial

location is set to be the corridor, where no contextual tutorial is activated. Once the list of

contextual tutorials isselected, itsdatastructure is transformedtobytesandsenttotheGUI

throughtheDLAConnectioncontextualTutorialsConn.

Whileaspecificrangeoftutorialsareservedtouserswhentheychangelocation,alltheother

tutorials are also available, since the personalized list of tutorials is sent through the DLA

Connection before the update of contextual tutorials is handled. The subsets of tutorials

selectedaccordingtotheavailablelocationsare:

Location:KitchenCategory:CookingandDomestics

Location:Living‐roomCategory:Therapy

Location:BedroomCategory:Dressing

Location:BathroomCategory:Grooming


13.RULE‐BASEDEXPERTSYSTEM

Thenecessityof introducingan intelligent layer on the systemariseswhenhuman reasoning

hastobesimulatedusingmedicalknowledgeinordertotakedecisions.Theexpertsystemwill

beresponsibleforselectingthelistoftutorialsaccordingtotheuser’sneeds.

Other AI methods, such as a Case‐Based Reasoning System (CBR), have been considered in

order to implement this expert system. However, the cost of its implementation was

determinedtobemuchhigherthanusingareasoningenginebasedinrules.

ThetypeofdeductivereasoningaRuleEngineusesisbasedonaknowledgebasewhererules

are activated as conditions have a positive evaluation, generating new facts as a result. In

contrast, CBR is amethodbased on experience,where a case base is created and extended

when the current case have no similarities with previous cases. This would be a complex

approachtoasolutionforthehumanreasoningthesystemsneedstosimulate.

13.1JESS

JESS5 is aRuleEngineand scriptingenvironmentwrittenentirely in JAVA languageby Ernest

Friedman‐Hill at Sandia National Laboratories. JESS allows to build software that has the

capacitytoreasonusingsuppliedknowledgeintheformofdeclarativerules.Oneofthemain

advantagesJESSoffersamonganotherRuleEngines is itssimplicitytobe integratedwithany

otherJAVAapplication,asitisMAS.

JESS has two relevant features concerning inference rules: it uses a more efficient and fast

patternmatchingalgorithm(theRetealgorithm),andBackwardschaining,amethodthatstarts

withalistofgoalsandworksbackwardstoseeifthereisdatawhichwillallow ittoconclude

anyofthesegoals.

5http://www.jessrules.com/


13.2KnowledgeBase

A Rule‐Based Expert System is made up of many such inference rules. They are entered as

separate rules and it is the inference engine that uses them together to draw conclusions.

Medicalknowledgehasbeenusedtostablishtheknowledgebaseforthisreasoningprocess.

Doctors and geriatricians supervising the SHARE‐it project have determined, for each of the

existingtutorials,thegradeofseverityallowedforthoseconditionsthatdefinetheuserprofile.

Thetwotablesbelowillustratethisclassification.

Regardingthepossiblegrades, fourvaluesaregenerallyusedforevaluatingseverityofvisual

impairent, sound impairment, fine coordination and dementia: 0 ‐ meaning the user is not

affected,1 ‐mildgrade,2 ‐moderategrade,and3 ‐ severegrade.However, severegrade is

only considered for sound impairment. Regarding the rest of the conditions, possible values

are:0–notaffected,1–affected,foraphasia;0‐none1‐left(leftside),2‐right(rightside)

fordementia.

PHYSICALCONDITIONS

TutorialVisual

impairmentSound

impairmentFinecoordination

Washingthedishes 0,1 0,1,2 0,1

Washingclothesbyhand 0,1,2 0,1,2 0,1

Goingoutdoors 0,1,2 0,1,2 0,1,2

Puttingshoeson 0,1,2 0,1,2,3 0,1,2

Preparingcoffee 0,1,2 0,1,2 0,1

Cookingpasta(Simple) 0,1 0,1,2 0,1

Cookingpasta(Complex) 0,1 0,1,2 0

Preparingcoldsoup 0,1 0,1,2 0

Makingasandwich 0,1,2 0,1,2 0,1,2

Takingdrugs 0,1,2 0,1,2,3 0,1

Morninggrooming 0,1 0,1,2,3 0,1

Brushingteeth 0,1,2 0,1,2,3 0,1


COGNITIVECONDITIONS

Althoughthereisanextenserangeoftutorials,someofthemhavenotbeenconsideredforthe

implementationoftheRuleEngine,sincetheywerenotcharacteristicenoughtodifferenciatea

group of users. However, the implementation fulfills those requeriments related to scabilty,

allowingtheservicetobeextendedeasily.

Tutorialsconsideredfortheimplementationandtestingphaseare:Goingoutdoors,Preparing

Coffee,Cookingpasta(simple),Cookingpasta(complex),TakingdrugsandMorninggrooming.

Twodifferentcookingpastatutorialswereconsideredwhencreatingtheknowledgebasefor

theRuleEngine,dependingontheusergeneraldegreeofimpairment:mild(complextutorial)

or moderate (simple tutorial). Complex cooking pasta includes holding and carrying heavy

kitchenwarewhilesimpledoesnot.

Tutorial Aphasia Neglect Dementia

Washingthedishes 0 0,1 0,1

Washingclothesbyhand 0 0,1 0,1

Goingoutdoors 0,1 0,1,2 0,1

Puttingshoeson 0,1 0,1,2 0,1,2

Preparingcoffee 0 0,1,2 0,1

Cookingpasta(Simple) 0 0,1,2 0,1

Cookingpasta(Complex) 0 0 0

Preparingcoldsoup 0 0,1 0,1

Makingasandwich 0,1 0,1,2 0,1

Takingdrugs 0,1 0,1,2 0,1,2

Morninggrooming 0,1 0,1,2 0,1,2

Brushingteeth 0,1 0,1,2 0,1,2


13.3WorkingMemory

The working memory containing the knowledge repository is composed by 7 facts, whose

contentandrelationsaredefinedonatemplate.

‐ One fact for theuser profile,where every slot corresponds to one of the conditions:

visualimpairment,soundimpairment,finecoordination,aphasia,neglect,dementia.

‐ Six factscorrespondingtotheavailabletutorials.Eachofthesefactshastwoslots: its

nameandaBooleanvalueindicatingifthetutorialisselectedornot.

IntheexecutionoftheRuleEngineseveralfilesareinvolved:RuleEnginePlan.java.Thisclassdeclaresan instanceoftheRuleEngineandresettheworking

memory to assert the initial fact. It calls the UserProfile.clp file, containing the templates,

queries, factsandrulesdefinition.Onceallthesestructuresaredefined,thecurrentvaluesof

theuserprofilefactareretrievedfromtheBeliefBaseofthePatientAgent,andassertedinto

theworkingmemory.Following,theRuleEngineisruntoasserttherestoffactsandexecute

therules.

Finally,aqueryisusedtoretrievethegroupoftutorialsselectedbytheRuleEngineaccording

totheirBooleanvalue.Resultsobtainedareusedtoupdatethecorrespondingbeliefsonthe

BeliefBase.

UserProfile.clp. This is a JESS file containing all the templates, queries, facts assertions and

rulesdefinitions.

Each of the rules created for the inference reasoningcorrespond to one of the tutorialsand

followsthesamepattern:thecurrentvaluesoftheuserprofileconditionsarecomparedwith

those indicated in the knowledge base. If the condition is evaluated positively, the fact

correspondingtothattutorialintheworkingmemoryismodified,settinghisBooleanvalueto

true.Awarningmessageisalsoprintedout.


14.MEDIARECORDING

Videoclips used for illustrating each of the available tutorialswere recorded in CasaAgevole

andintheFondazioneSantaLuciafacilities(adaptedbathroomsforpatientsandoccupational

therapyarea)duringthesummer‐autumn2009inRome.TheITspecificationinsection8.2was

usedtorecordthefootages,followingthestepscomposingeachtutorial.

Medicalassesmentwasextremelyvaluableinordertofiximportantissuesregardingthevideo

recording:whatwasthebestangleforfilming,whichwastheproperordertopresenttheitems

needed,howtheactorshouldactinfrontofthecamara,etc.

Additionally, several changes were introduced while recording, since medical considerations

such as including a new cooking pasta tutorial for those users not able to carry a pot or

generalizing themorning grooming tutorial to enlarge its target populationwere taking into

consideration.

Audio samples in italianwere recorded voice‐over byDr.Nadia Tini and added to videoclips

during the experimentation phase, as a significant improvement to get users focused on the

instructionsgiven.

Figure18.RecordingoftheGroomingandGoingOutdoorstutorials


Figure19.RecordingofthePreparingCoffeeandPreparingPastatutorials


TESTING_________________________________________________

Performanceengineeringistheprocessbywhichsoftwareandhardwarearetestedandtuned

with the intent of realizing the requiredperformance.The testing casesof the IT serviceare

included within the testing phase of SHARE‐it project, which execute scenarios 1, 2 and 36.

These scenarios have put the different technologies and integrated services (among which

thereisITservice)tothetestwiththeaidof15italianvolunteers.

We are using some of those concepts to analyze the following three aspects of SHARE‐it

architecture’squalityofservice[23]:

• Performance(responsetimes)

• Scalability(throughput)

• Reliability(availabilityandfunctionalintegrity)

Withoutastrategy,performanceengineeringissimplyanexerciseintrialanderror.

15.TESTCASES

15.1TestingCycles

Designing test cases can be time consuming in a testing schedule, but they areworth giving

timebecausetheycanreallyavoidunnecessaryretestingordebuggingorat least lower it. In

our case, engaging volunteers with disabled conditions in testing ITwith real activities is a

guaranteedwayofensuringthattheserviceisusable,effectiveandconsistent.

6Refertosection8.1forafurtherexplanationofIT’sScenarios


15.1.1PlanofAction

When getting prepared for an on‐site demonstration it is necessary to plan in advance the

followingpoints:

i. Pre‐Test,self‐assessmentchecklist

ii. Haveallnecessarysoftwareandhardwareatworksite

iii. TestalljobfunctionswithAssistiveTechnologyrunning

iv. Reconfiguresettingsandconfigurationsifnecessary

v. Developedscriptsforusability

(i)wasdevisedasapre‐testwherethemedicalteamassessedthetechnologiesinvolvedinthe

testcase–calledscenario—astoconfirmthateachelementoftheSHARE‐itarchitecturewas

compliantwiththespecificationandthattheintegrationofall involvedelements(includingIT

service)wasalsosafeandconsistent.Thispre‐testalsocovered(ii)and(iii).

Figure20.DeploymentofsensornetworkinCasaAgevole


The test site for SHARE‐itwasCasaAgevole,a pre‐existing facility in Fondazione Santa Lucia

thatwasenhancedwiththedeploymentofanetworkofsensorsusedforthecontextualization

oftheITservice.

When it was necessary, the medical team indicated some elements to be modified and/or

reconfiguredbeforetherealtestwiththevolunteersinordertoaddresspoint(iv).

15.1.2Experimentation

BeforeaddressingthetecnicalevaluationoftheSHARE‐itelementsinteractingwithITinthe3

scenariosusedat theexperimentationphase, it is important to clarify some issues regarding

thewholeprocess.

On the one hand, the decision to deploy thewhole architecture in CasaAgevole in order to

approachinasmuchaspossiblearealhouse.Takingintoconsiderationthatallthefacilitiesat

Casa Agevole are contained in a 60m2, it is easy to understand that mobility tended to be

challengingwhenusingroboticplatformswherethetouchscreenwasintegrated.

Ontheotherhand,anotheraspectto consider is thedynamicsoftheexperimentation,since

theservicepresentedtothefirstvolunteer(ID01)andtheonepresentedtothelastvolunteer

significantlydiffers.Newpossibilitiesofinteractionorrefinementsappearedinthefirsttesting

and they were implemented in situ. This method was very instructive and interesting but

modified the perception that volunteers had about the system and, therefore, users’

perceptionwasbetterattheendfotheexperimentationphase.

Regarding the teams participating in the evalutaion of the testing, there was a cluster of 7

peoplepermanently inCasaAgevolesupportingexperimentsandevaluatingtheprocess.This

constantrevisionassuredaverystableevaluationcriteriaduringallthetestingphase.


16.OUTCOMES

16.1Tutorials’ExperimentationResults

Inthissection,theresultsfromthethreeselectedscenariostestingtheITservicearedescribed,

alongwiththosechangesrequestedandimplementedforeachscenario.

16.1.1Scenario1:Albertotakeshisdrugs

Thefirstscenariotestedontheexperimentationstageentailedlaunchingareminderfollowed

bythecorrespondingtutorial.Someofthecomponentswerenotavailableatthatmoment(like

some domotic sensors), although it did not affect the functionality of the IT service and its

connectiontoothers.ITusabilitystronglydependedontheproperoperationoftheGUI,asthe

therapy tutorialwas offered through it.Being the first time tutorialswere performedby the

users,considerablechangesweremadeuntiltheystartedtofittheuserandmedicalrequests.

FunctionalEvaluation:Thefollowingtablesynthesizesthefunctionalevaluationofthetherapy

tutorialwithinthecontextofscenario1.TheTaskExecutioncolumnsevaluatetheexecutionof

thetaskdefinedbytheservicewhiletheServiceSupportevaluatethedegreeofsupportthat

thisservicesupposedtothetestuser.Ascalefrom0to7wasused,being7thebestscore.

TaskExecutionResults ServiceSupportResults

VOLUNTEERID VOLUNTEERIDSERVICE TASK

1 2 3 4 5 1 2 3 4 5

TutorialsTherapyTutorial

5 5 5 5 6 4 2 5 1 5


Someuserswerenotabletointeractsucessfullywiththetherapytutorialsincetheywerenot

toldabouthowthetutorial’sflowworked.Therefore,itwasagreedthatsomepre‐trainingwas

neededinordertoguidethem.

Regardingthetutorial’spresentation,severalchangeswereappliedtoitsfunctionalityinorder

to improveusability.Mostofthesechangeswererequestedbythemedicalteamwhenusers

encounteredseriousdifficultiesatperformingtheactivity.Themostsignificantoneswere:

• Introducing a forced delay between the tutorial‐steps in order to let the volunteers

perceivetheinformationafterafewsecondsfororientation.

• Removing the confirmation‐phase between each tutorial‐step, leaving a unique

confirmationattheendofthetutorial. Itwasproventhatthis functionalitytopursue

users’interactionwasactuallymoreconfusingthanhelpful.Insteadoftheconfirmation‐

screen,atimer(implementedbytheGUIdeveloper)wouldshowadialogueaskingusers

iftheywanttoproceedortorepeatastepwhennoactionwasperceivedafteracertain

periodoftime.

• Dividingthefirsttutorialstep(theoneaskingtheusertomakesurenoneoftheitemsto

perform the activityweremissing) into two different steps, since users tended to be

distractedbythelengthoftheinstructionduringthetesting.

• Audio support was requested to be added to the videoclips, as users react easier to

voiceinstructionsthantotextguidelines.

TechnicalEvaluation:thetherapytutorialoperatedrobustlyandquickly,beinghandledinless

than0.5seconds,anacceptableresponse.DLAconnectionswereprovedtobeworkingasthe

listoftutorialswasproperlyinitializedandthetherapytutorialwaslaunchedandservedtothe

GUIasexpected.

As the therapy tutorialhadoneof themain roles in the scenario,being the first time itwas

testedwith the target users, a lot of feedbackwasgathered. Some additional functionalities

affecting the technical implementation of IT service were required, such as substituting all

integratingsoundwithvideoforabettercomprehension.


Outcomes:Althoughvolunteersgetusedtothetutorialsworkflowbroadly,someencountered

difficultiesinteractingwithacomputer,astheywerenotaccustomedtonewtechnologies.This

factswere explicit by trying to issuevocal commandsafter finishing an step performance or

beingconfusedwiththeinteraction‐flow.However,afterafewtrialsandfurtherexplanations

theyeventuallybecamefamiliarwiththeapplication.Thisfactwasencouraging,confirmingIT

asagoodassistivetechnology.

Thisscenariowasveryprofitableintermsoffeedbackforcheckingthespecification(technical

andfunctional)oftheITserviceandimproveitspresentationtoend‐users.

INTELLIGENTTUTORIALS

VolunteerID

PresentActiveRole

StableReactionTime

ExecutionErrors

Comments

1 X X X 5 □Toomanyconfirmationstepsweredetectedinbetweentutorialsteps

2 X X X 5 □Maybesoundcouldbeintegratedtovideosforbetterunderstanding

3 X X X 6 □Theusergotusedtothetutorialsafterafew

trials

4 X X X 6 □

5 X X X 6 □


16.1.2Scenario2:Pietrogoestothechurch

Thisscenariorequestedtodisplayareminder,awayfortheusertoreachthecorresponding

tutorial.Aftertheexperienceoftestingthepreviousscenario, ITservicewasmoreconsistent

andusable,whichmadeitmoreappropiateforitstargetpopulation.

FunctionalEvaluation:The following table synthesizes the functionalevaluationof IT service

forscenario2.TheTaskExecutioncolumnsevaluatetheexecutionofthetaskdefinedbythe

servicewhiletheServiceSupportevaluatethedegreeofsupportthatthisservicesupposedto

thetestuser.Ascalefrom0to7wasused,being7thebestscore.



11 12 13 14 15 11 12 13 14 15

TutorialsGoingOutTutorial

6 6 6 6 6 6 6 2 6 6

On the one hand, the going out tutorial brought up a newsituation: guiding users onwhich

steps ought to be followedbefore leavinghomebut performed in different places, so users

werenotstaticduringthetutorialandhadtokepttheirconcentrationwhenmovingaround.In

allcases,mobilitywasnotperceivedasahindrancewhenfollowingthetutorial.

Ontheotherhand,usersweretoldhowthe interfaceworkedbeforestartingtoperformthe

activity,sothattheycouldbeabletorespondtothetutorial’sworkflowproperly.

Regardingthetutorial’spresentation,changespreviouslyappliedtoitsfunctionalityprovedto

beadequateinordertogetmoredynamicandsimpleguidelines.

Technical Evaluation: the going out tutorial operated robustly and run flawless as was

expected. The incoming tutorial was handled in less than 0.5 seconds, being an acceptable

response.


Audiosampleswererecordedandaddedtothevideoclipstomadetheunderstandingprocess

easier. Voice guidelines proved to be very effective in terms of focusing users on what the

intructionsaskthemtodo.

Thisscenariopresentedatutorialthathadtobefolloweddynamicallyondifferentplacesofthe

house,comparedtoothersthathaveall thestepsperformed inthesamespot.This factwas

notreflectednegativelyinthetechnicalexecution.

Outcomes: the IT service standed out to be one of the most effective one, providing clear

instructionsondailyactivities.Althoughtherewasanupgraded levelof complexityregarding

the user mobility in order to perform instructions, results obtained suggest that no extra

mental workload was added for the users. Changes previously applied to the tutorials’

presentationprovedtobepositive,sinceusers’reluctancetowardtheapplicationwasreduced.


VolunteerID

PresentActiveRole

StableReactionTime

ExecutionErrors

Comments

11 X X X 6 □

12 X X X 6 □

13 X X X 6 □Theuserdidnotwantto

followthetutorial.

14 X X X 6 □

15 X X X 6 □


16.1.3Scenario3:Francescacookspasta

The focus of this testing phasewas to consolidate the results from previous phases and to

integrate the now available US localization system (sensors) in the scenario execution.

Localizationwascrucialinthisscenarioinordertodetectwhentheuserenteredinthekitchen

inordertoofferhimthecontextualtutorial.

FunctionalEvaluation:The following tablesynthesizes the functionalevaluationof IT service

forscenario3.TheTaskExecutioncolumnsevaluatetheexecutionofthetaskdefinedbythe

servicewhiletheServiceSupportevaluatethedegreeofsupportthatthisservicesupposedto

thetestuser.Ascalefrom0to7wasused,being7thebestscore.



16 17 18 19 20 16 17 18 19 20

TutorialsCookingPastaTutorial7

6 6 6 6 6 6 6 4 6 6

After two scenarios tested, tutorials’ presentation were muchmore suitable for the target

users,speciallywhentestingwithvolunteerssufferingfromamoderatedisability.Ingeneral,

theservicewasperceivedasveryusefulinsupportingend‐users.

Cookingscenarioalsopresentedtwodifferentcookingtutorialsdependingontheuserprofile:

maleusersfollowedapreparingcoffeetutorialwhilefemalesfollowedacookingpastatutorial.

Theculturalbackgroundandageoftheusersleadthisdistinction,asmaleuserswerenotused

tocookfood.

7MalevolunteersperformedPreparingCoffeeTutorial


Technical Evaluation: the last scenario did profit from the collected experiences from the

previous test‐runs. After implementing the numerous changes and suggestions from the

medicalpartners,ITservicewasabletosuitmostoftherequiredneeds.

TheUSlocalizationsystemworkedproperlyinordertoofferacontextualtutorial,provingthat

theintegrationbetweenthedifferentagentswasprofitable.

Thecookingtutorialoperatedrobustlyandrunflawlessaswasexpected.Theincomingtutorial

was handled in less than 0.7 seconds due to the localization process, being an acceptable

response.

Outcomes:thecookingtutorialwasperformedinCasaAgevolekitchenusingrealkitchenware

that was the same shown in the recorded videos, in order to make easier for users having

cognitive disabilities to recognize the tools to use. In addition, there was a certain level of

complexityonperformingsuchatutorial,asitinvolveddifferentskillsandgoodcoordinationin

ordertoperformtheactivityasexpected.

Contextual tutorials were also proven to work properly, being displayed when the system

acknowledgedthatuserswhenchangingtheirlocation.


VolunteerID

PresentActiveRole

StableReactionTime

ExecutionErrors

Comments

16 X X X 6 □

17 X X X 6 □

18 X X X 6 □Theyweretriggered

manually

19 X X X 6 □

20 X X X 7 □


CONCLUSIONS____________________________________________

17.TEMPORALANDECONOMICANALYSIS

17.1TemporalAnalysis

Theprojectphasesincludedinthetemporalanalysisare:

• Specification.ThisphaseincludestheintroductiontotheSHARE‐itproject(5days),the

researchonthetechnologiestouse(28days),theStateofArtdescription(8days),and

theanalysisofthesystemintermsofrequirementsandusecases(6days).Alsoarisk

analysis(2days)isincludedintheproject,asdeviationsarelikelytooccur.

Considering that during this early phase of the project the time dedicated was

approximately4hoursperday,thesumoftimeconsumedforthespecificationwas:

49daysx4hours/day=196hours

• Design. Thisphase includesthe interactionworkflowdesign(4days)andthecontent

description(7days),referringtotheADLsandscenariosdesign.

Consideringthatduringthisphaseoftheprojectthetimededicatedwasapproximately

8hoursperday,thesumoftimeconsumedforthedesignwas:


• Implementation.Thisphase includesthecodingofthe IT’s functionalities (28days)as

wellastheimplementationoftheIntelligentLayerusingaRule‐BasedExpertSystem(8

days).


6hoursperday,thesumoftimeconsumedfortheimplementationwas:



Forthevideorecording(4days),atechnicalexpertwashired,working4hours/day:


Figure21.Temporalanalysisschema

• Testing. This phase includes all the experiments tested in Casa Agevole (17 days),

considering time dedicated to correct functional errors and implement new changes

requestedbythetechnicalandmedicalteam.


8hoursperday,thesumoftimeconsumedforthetestingwas:


• Documentation.Thewritingofthereportstartedthefirstday,includingthemotivation

and the introduction to the technologies. However, most of the documentation was

reviewedafterthetestingphase.



4hoursperday,thesumoftimeconsumedforthedocumentationwas:


Anapproximationtothetotalamountofhoursdedicatedtotheprojectresultsfromthesumof

hourseveryphasetook:

196+88+216+16+136+196=848hours

17.2EconomicAnalysis

Therearethreeimportantaspectstotakeintoconsiderationwhendoinganeconomicanalysis:

• Hardware.Applicableexpensestothedevelopmentplatformoftheapplication.

• Software.Applicableexpensesfromtheacquisitionofthosecomputertoolsusedduring

thedifferentphasesoftheproject.

• HumanResources(HR).Applicableexpensestothosetaskscarriedoutasaresearcher,

analystorprogrammer.

Hardwareexpenses.ConsideringthattheanalysisoftheITservicesolution,thedevelopment

of the application and the individual testing have been carried out in the same computer,

expensesarisesfromthecostofthefollowingdevice:

Quantity Description Price

1PCIntelCoreDuoT75002.22GHz

4GBRAMMemory,120GBHardDisk700€

TOTAL 700€


The computer used during all the project, whose characteristics are described in the table

above,waspowerfulenoughtofulfillnon‐functionalrequirementsrelatedtooperability,such

asfastdisplayofcontentsanddataintegrity.

Inthiscostbreakdown,thepurchaseofavideocamera isnot included,asatechnicalexpert

washiredforthevideorecordings.

Softwareexpenses.Computertoolsusedduringthedifferentphasesoftheprojectare listed

below. Some sofware components of the SHARE‐it project has been used during the testing

phase,suchastheGUIortheDLAserver.

Description Price

EclipseClassic3.5.1 0€

JavaSEDevelopmentKit6 0€

JADEXBDIAgentSystemv2.0 0€

MicrosoftWord2007 150€

JESSVersion7.1p2 0€

SubversionclientTortoiseSVN 0€

WindowsMovieMakerVideoEditing 0€

GanttProjectTool 0€

TOTAL 150€

Itwasintendedtouseasmanyopensourcetoolsaspossibleduringthediferentphasesofthe

project.Therefore,sotwareexpensesarisesfromtheMSOfficepackage’slicenseforusingMS

Word2007.Therestofthesoftwarecanbedownloadedforfreeandinstalledeasily.


HumanResourcesexpenses.Thetimeinvestedintheprojectisdistributedamongthedifferent

phasesasdetailedintheprevioussection,thetemporalanalysis.

Consideringthataanalyst,aprogrammerandamediatechnicalexpertareneededinorderto

coverthedifferentHRneedsoftheproject,expensesbreakdownisestimatedasfollows:

Thetotaleconomiccostoftheprojectresultfromthesumofthehardware,softwareandHR

expensesanalyzed:

TypeofCost Cost

HardwareExpenses 700€

SoftwareExpenses 150€

HumanResourcesExpenses 29160€

TOTAL 30010€

Individual DevelopmentStage Hours Cost/hour Cost

AnalystSpecification+Design+

Documentation380h 40€ 15200€

ProgrammerImplementation+Testing+

Documentation452h 30€ 13560€

MediaTechnicalExpert

VideoRecording 16h 25€ 400€

TOTAL 848h 29160€


18.CONCLUSIONS

Currently,elderlypeoplewithcognitiveandphysicalimpairmentusuallyendupinhospitalsand

geriatricinstitutions,withthesubsequenthighexpensestheirrelativesaccumulate.Moreover,

it is proven that a familiar environment facilitates the patients’ enhancement and their

willingnesstobecomemoreindependentfromcaregivers.

SomeofthemostrelevantgoalsthatITpursueare,ontheonehand,toincreasethestandard

ofcarebyaugmentingtheuser’sautonomy,andontheotherhand,tomakecaregivers’work

easierandmoreprofitable.Havingthesegoalsinmind,itisrighttoassertthatthesolutionthat

ITserviceoffers isaproofofconceptofthebeneficial impactthatnextgenerationofATwill

haveinelders’everydayroutines.

TheSHARE‐itexperimentationphaserepresentedanimportantsourceofnewinformationand

knowledge for the improvement of the IT application. Testing scenarios allowed not only to

provethatsimpleuseofsoftwarefacilitatestheintegrationofATintothecareprocess,butalso

toreaffirmthatneitherexternalsupportisdirectlyneedednorpriortechnologicalexperienceis

required for the reinforceofusers’ autonomy.The test site choosen for theexperimentation

phase‐CasaAgevole‐allowedatightcontrolofdecisionontheenvironmentwithoutanyloss

ofrealism.

The testing phase also led to identify that part of the IT success heavily depends on the

personalization, since different personswith different disability profiles can have a different

approachtotheirautonomyaccomplishment.

18.1Futurelinesofresearch

Considering the outcomes obtained after testing IT, several improvements can be applied to

theserviceasafuturework.

Firstly,toachieveamorecomplexpersonalizationbasedonanextendeduserprofile.Themore

accurate the interaction would be, the better approach to the user’s autonomy: too much


supportcouldbeasnegativeastoolittle,underminingresidualcapabilitiesorjuststressingthe

user.

Secondly,anextensionoftheavailabletutorialswouldbeveryprofitable,addingnewroutine

activitieswhoseprocedural tutorialmighthelptheusertobecomemore independent.These

activitiescouldbefocusedondifferenthoursalongthedayforamorefunctionalservice.

Thirdly, to expand the target population by aiming the IT service at general users requiring

occupational therapy for cognitive enhancement.Most of the activities this type of subjects

performsconsistonveryconcreteandsimpleactions.Therefore,thetutorials’designwouldbe

adaptedtoanewsetofrequirements.

Lastly, to integrate the IT service with one of the other services that the MAS offers: the

BiometricCognitiveModule.Themaingoalwouldbetotrackusersbymeansofthebiomedical

measuresinordertoknowhowtheyreactwhenperformingtutorials(e.g.theuseristoomuch

stressedorinthecontrary,s/heinagoodmood).Retrievingthisinformation,newgoalscanbe

pursuedbycreatingdifferentflowsofinteractionincasetheuserpresentsacertainemotional

state.

Wearenowattheverybeginningofanewresearchregardingassistiveintelligenttools,asthe

presented model for tutoring users with disabilities has been enhanced by very fruitful

interaction with real patients and, therefore, the system has the necessary inputs to be

improvedandenriched.Additionalchallengeswillrequirecollaborationwithcolleagueshaving

expertiseinmachinelearning,naturallanguageprocessing,roboticsandmachinevision.


APPENDIX_______________________________________________

A1.GLOSSARY

A1.1Medicalterms

(byalphabeticalorder)

ALZHEIMER. Themost common form of dementia. In the early stages, the most commonly

recognizedsymptomismemory loss,suchasdifficulty inrememberingrecently learnedfacts.

As the disease advances, symptoms include confusion, irritability and aggression, language

breakdown,long‐termmemoryloss,andthegeneralwithdrawalofthesuffererastheirsenses

decline.Gradually,bodilyfunctionsarelost,ultimatelyleadingtodeath.

APHASIA. Language disorder which is presented when there is difficulty in using or

understanding spoken andwritten language or is completely lost the ability to do either. A

personwithaphasiamaythinknormally,sousingbabytalkisinappropriate.

APRAXIA.Neurological disorder characterized by loss of the ability to execute or carry out

skilledmovementsandgestures,despitehavingthedesireandthephysicalabilitytoperform

them.Apraxiamaybeacoompaniedbyalanguagedisorder(aphasia).

AUDITORYAGNOSIA.Inabilitytorecognizesounds.

CAREGIVER.Personalwhogivesmedicalassistanceandcaringsupporttotheuserathome.

COGNITIVE IMPAIRMENT. Capacities of the nervous system are limited or impaired with

difficultiesexhibitedinoneormoreofthefollowingareas:useofmemory,controlanduseof

cognitive functioning, sensory and motor skills, speech, language, organizational skills,

informationprocessing,affect,socialskills,orbasiclifefunctions.

DALTONIC.Inabilitytodistinguishredfromgreen.

DEMENTIA. Progressive decline in cognitive functions affectingmemory, attention, language

andproblemsolving.Highermentalfunctionsareaffectedfirstintheprocess.Especiallyinthe

laterstagesofthecondition,affectedpersonsmaybedisoriented intime(notknowingwhat


dayoftheweek,dayofthemonth,orevenwhatyearitis), inplace(notknowingwherethey

are),andinperson(notknowingwhotheyareorothersaroundthem).

DYSPHAGIA. Sensation that suggests difficulty in the passage of solids or liquids from the

mouthtothestomach.

HEMIPARESIS.Paralysisaffectingonlyonesideofthebody.

HIPFRACTURE.Abreakofthetoppartofthefemurbonewhereitconnectstothepelvis.Only

asmallportionofpatientsretaintheirpreviousmobility,whileabout20%willrequirenursing

homecare.

IPOVISION.Visualacuitybelow1/10.

MIXEDIMPAIRMENT.Simultaneouspresenceofmotorandcognitivedisabilityonapatient.

NEGLECT.Unilateralspatialneglectisaconditionwhichreducesaperson’sabilitytolook,listen

ormakemovementsinonehalfoftheirenvironment.Thiscanaffecttheirabilitytocarryout

manyeverydaytaskssuchaseating,readingandgettingdressed.

PARKINSON.Itischaracterizedbymusclerigidity,tremor,aslowingofphysicalmovementand,

in extreme cases, a loss of physicalmovement. Secondary symptomsmay include high level

cognitivedysfunctionandsubtlelanguageproblems.

PATIENT.Individualwithcognitiveormixedimpairment(cognitiveandphysical).Regardingthe

system,usuallyisknownas‘user’.

PRESBYOPIA.Conditionwhere theeyeexhibitsaprogressivelydiminishedability to focuson

nearobjectswithage.

STROKE.Itoccurswhenthebraindoesnotgetsufficientoxygen.Damageontherightsideof

thebrainmayimpairmovementandsensationontheleftsideofthebody.Damageontheleft

side of the brainmay affectmovement on the right side. Speech and language, vision loss,

breathing,swallowing,balance,hearing,andbladderandbowelfunctionmaybeaffected.

VISUALAGNOSIA.Lossofabilitytorecognizeobjects,facesandwords.


A1.2Technicalterms

(byalphabeticalorder)

ADLs(ActivitiesofDailyLiving).Basictasksapersonneedstocarryoutdaily,suchasdressing,

cookingandtoileting.

AT.AssistiveTechnologies.

DLA(DistributedandLayeredArchitecture).DLAArchitecture isusedtodevelopcooperative

systemswithdifferentinteractingmodulesdistributedalongseveralmachines.Itimplementsa

distributedsharedmemorymodel,withacentralelementthatmanagestheaccessandstorage

oftheshareddatafromtheothermodules.

ENVIRONMENT AGENT. Its basic target is to distribute the information from all available

sensorstoalltheagentsinterested.

GUI.GraphicalUserInterface.

HOMEAGENT.Representsthehomeentity,whichwillbesituatedinahomeserver.Amongits

objectives are tomaintain themonitoring of the users, tomanage their daily living activities

andtheirprofiles.

IT.IntelligentTuturials.

i‐WALKER.Assistive devicewith four conventional wheels and two degrees of freedom that

improveuser’sstabilityandsafetywhilewalking.Itmaysupportupto50%oftheuser’sbody

weight,beingidealforweakkneesoranklesorseverebalanceproblems.TheintelligentWalker

hastwohandlesthattheuserholdswithbothhandsforinteraction.

MAS(Multi‐AgentSystem).Smallsoftwareentities–softwareagents–withspecialcapabilities

(autonomous, reactive, pro‐active and social) are used instead to interact in a flexible and

dynamicwaytosolveproblemsmoreefficiently.

PATIENT AGENT. An instance of this agent should provide all the available and permitted

services to each user, such as security,mobility,monitoring and help services, or those that

couldbeaddedinthefuture.Patient‐agentsrunsinPDAsorUltra‐MobilePCs.


ROLLAND.Semi‐autonomouswheelchairwhichisabletofunctionsafelyindoorsandoutdoors

inappropriatelyset‐upenvironments.

SHARE‐it.SupportedHumanAutonomyforRecoveryandEnhancementofcognitiveandmotor

abilitiesusing InformationTechnologies.Europeanprojectwhich includesADLtutorasoneof

itsmultipleservice.

WI‐FI.Wireless networking technology that uses radiowaves to providewireless high‐speed

Internetandnetworksconnections.


A2.STORYBOARDS

TheGUIthathasbeenimplementedbyanotherdeveloper,fulfillsthefunctionalrequirements

and use cases of the IT specification for the tutorials’ display. Workflows defined when

performingatutorialshavebeenalsoconsidered.Medicalassesmenthasbeeninvolvedduring

thedesign,inordertodevelopasuitableinterfacewhosetargetpopulationiselderlypeople.

Figure22.Listoftutorials(categories)‐screenshot


Figure23.Firststepoftutorials‐screenshot

Figure24.Laststepoftutorials‐screenshot


Figure25.Requestforhelp/Reportanalert‐screenshot


REFERENCES________________________________________

1.[Camarinha‐Matos,1999]“Intelligentmobileagentsinelderlycare”,Camarinha‐MatosL.M,

VieiraW.RoboticsandautonomoussystemsJournal,vol.27,nº1‐2(pp57‐75).

2.[Zukerman,2003]“EldercareforDummies”,RachelleZukerman,Ph.D.WileyPublishing,

2003.

3.[Bellifemine,2007]“Developingmulti‐agentsystemswithJADE”,FabioBellifemine,Giovanni

Caire,DominicGreenwood.WileySeries,2007.

4.[Mihailidis,2007]“TheUseofanIntelligentPromptingSystemforPeoplewithDementia”,

AlexMihailidis,JenniferBoger,MarcelleCanido,JesseHoey.DesigningforSeniorsPublication–

specialedition.Interactions(July+August2007).UniversityofToronto,UniversityofBritish

Columbia.

5.[Carmien,2002]“Socio‐TechnicalEnvironmentsSupportingDistributedCognitionfor

PersonswithCognitiveDisabilities”,StefanParryCarmien.MasterDissertation,Universityof

Colorado,2002.

6.[Nkambou,2001]“PlanningAgentsinaMulti‐agentsIntelligentTutoringSystem”,Roger

NkambouandFrodualdKabanza.Proc.of14thInternationalConferenceonIndustrial&

EngineeringApplicationsofArtificialIntelligence&ExpertSystems.LectureNotesinComputer

Science.pp:921‐930,2001.

7.[Olivier,2007]“Designingtechnologyforpeoplewithdementia”,S.CLindsay,P.Olivier,

A.L.RobinsonandA.Monk.WorkshopontheFamilyandCommunicationTechnologies,

Newcastle,UK,2007.

8.[Freedman,2001]“AdaptiveProcessinginaMedicalIntelligentTutoringSystem”,Reva

Freedman.WorkshoponAdaptationinDialogueSystems,NAACL2001,Pittsburgh.

9.[Freedman,2000]“WhatisanIntelligentTutoringSystem?”,RemaFreedman.Finaldraft.

PublishedinIntelligence11(3):15–16(Fall2000).NorthernIllinoisUniversity.


10.[Pokahr,2008]“Jadex‐EngineeringGoal‐OrientedAgentsin:DevelopingMulti‐Agent

SystemswithJADE”,AlexanderPokahr,LarsBraubach,AndrzejWalczak,WinfriedLamersdorf.

11.[Pokahr,2007]“AnArchitectureandFrameworkforAgent‐BasedWebApplications”,

AlexanderPokahr,LarsBraubach.5thInternationalCentralandEasternEuropeanConference

onMulti‐AgentSystems(CEEMAS2007).

12.[Julián,2000]“AgentesInteligentes:elsiguientepasoenlaInteligenciaArtificial”,V.Julián,

V.Botti.SoftwareDepartment,ValenciaTechnicalUniversity.May‐Jun2000,ATI2000.

13.[SHARE‐it,2008]ISTSHARE‐itProjectD5.1b“Scenariosdefinitionandperformancecriteria

(FINAL)”.VersionF.

14.[SHARE‐it,2008]ISTSHARE‐itProjectD3.1“AgentBasedControlLayout:StateoftheArt.

Architectureproposal.Distributionofthecontrolschemeintheenvironmentandthemobile

entities”.Versionv1.0C.

15.[SHARE‐it,2007]ISTSHARE‐itProjectD2.1“Definitionoftheuserandactivity,tasksand

activitiesofdailyliving”.Version3.0.

16.[Hoey,Poupart,2005]“POMDPModelsforAssistiveTechnology”,InProceedingsofthe

AAAIFallSymposiumonCaringMachines:AIinEldercare,Stanford,2005.Dept.ofComputer

Science,UniversityofToronto.

17.[Mostow,1999]“ProjectLISTEN’sReadingTutor:InteractiveEventDescription”,Frontiers

inArtificialIntelligenceandApplications;Vol.158.RoboticsInstitute,CarnegieMellon

University,Pittsburgh(USA).

18.[SHARE‐it,2008]ISTSHARE‐itProjectD2.3“InteractionDesignReportandMiddleware”.

Version2.0.

19.[SHARE‐it,2009]ISTSHARE‐itProjectD2.4“InterfaceSpecification”.VersionF.

20.[SHARE‐it,2009]ISTSHARE‐itProjectD3.4“MobileAgent‐Basedcontrolarchitecture:In‐

depthtechnicaldescriptionoftheintelligentcontrolarchitecturerelatedtothemobile

entities”.


21.[COACH,2009]“CognitiveOrthosisforassistingACtivitiesintheHome”.

http://fast.isledev.co.uk/research/projview.php?id=1539

22.[SHARE‐it,2007]Contractfor:SpecificTargetedResearchorInnovationProject.AnnexI‐

“DescriptionofWork”.

23.[SHARE‐it,2009]ISTSHARE‐itProjectD5.4“ReportonPerformanceTestingofthePilot

Application”.FinalVersion.

24.[SHARE‐it,2009]ISTSHARE‐itProjectD5.5“FinalreviewandassesmentofSHARE‐it

componentsandtheirperformance”.Version12.

25.[Bratman,1987]“Intention,Plans,andPracticalReason”,M.Bratman.HarvardUniversity

Press.Cambridge,MA,USA.1987.

26.[RaoandGeorgeff,1995]“BDIAgents:fromtheorytopractice”,A.RaoandM.Georgeff.

ProceedingsoftheFirstInternationalConferenceonMulti‐AgentSystems(ICMAS'95).TheMIT

Press.Cambridge,MA,USA.1995.pp.312‐319.

27.[SusanBull,2004]“SupportingLearningwithOpenLearnerModels”.Universityof

Birmingham.4thHellenicConferencewithInternationalParticipation:Informationand

CommunicationTechnologiesinEducation,Athens,2004.

28.[Demirkan,2007]“Housingfortheagingpopulation”.EuropeanReviewofAgingand

PhysicalActivity.SpringerBerlin,2007.pp.33‐38.

29.[Abraham,2005]“Rule‐basedExpertSystems”,HandbookofMeasuringSystemDesign,

chapter103.AjithAbraham,OklahomaStateUniversity,Stillwater,OK,USA.(Extractedfrom

Wikipedianotations).

30.[Campana,2008]“FromChronicDiseasestoGeriatricSyndromes:ManagingComplexityin

K4CARE”,StateofArt‐FP6SpecificTargetedResearchorInnovationProjectThematicPriority

2:”InformationSocietyTechnology”.CAD,Rome(Italy).

31.[Pollack,2005]“IntelligentTechnologyforanAgingPopulation”,theUseofAItoAssist

elderswithCognitiveImpairment.AIMagazine,Volume26,Number2.


32.[Hayes‐Roth,1995]“Anarchitectureforadaptiveintelligentsystems”,Artificial

Intelligencepublication,Volume72,Issues1‐2(pp.329.265).KnowledgeSystemsLaboratory,

StanfordUniversity.

33.[SHARE‐it,2009]ISTSHARE‐itProjectD3.5“MobileAgent‐BasedControlarchitecture

software:Testingofthearchitectureabilitytoprovideefficientemergentbehavioursinreal

timeresults”.

Documents

SHAREit Tutorials Closed - UPC Universitat Politècnica de