erstolet

DesigningPersonalTele-embodimentEricPaulosJohnCannypaulos@cs.
berkeley.
edujfc@cs.
berkeley.
eduDepartmentofElectricalEngineeringandComputerScienceUniversityofCalifornia,BerkeleyAbstractAttheintersectionoftele-robotics,computernetworking,andhumansocialinteractionwehavechosentoexploreanareaweidentifyaspersonaltele-embodiment.
Atthecoreofthisresearchisanemphasisontheindividualpersonratherthantheintricatecomplexitiesofthemachine.
Whilethemechanicalelementsofoursystemareessentialtoitsover-allfunctionality,ourresearchisdrivensolelybythestudyandunderstandingofthesocialandpsychologicalaspectsofextendedhuman-humaninteractionsratherthanthelat-esttechno-gadgetry.
Inthispaperweemphasizetheimpor-tanceofthehumancomponentanddescribethedevelop-mentofonesuchsimple,inexpensive,internet-controlled,untetheredtele-robotorPRoP(PersonalRovingPresence)thatprovidesseveralfundamentalelementsofpersonaltele-embodiment.
1IntroductionOnlyafewdecadesagocomputerswerebeingpraisedsolelyontheirabilitytotacklecomplexmathematicalproblemswithlittlediscussionoffutureapplicationsbeyondtheirthenuseassophisticatedmilitaryandresearchlaboratorycalcu-latingengines.
Clearly,thecomputersoftodayhaveevolvedandassimilatedthemselvesintothedailylivesofcount-lesspeopleinwaysthatwereneverimagined.
Similarly,roboticsresearchoverthelastfewdecadeshaswitnessedamyriadofrevelingcontributionstoscienceandsociety.
Whilegivingproperpraisetothesecontributions,wepro-poseanaugmentationtocurrentroboticsresearchthatmayresultintheextensionofroboticsintothelivesofordinarypeopleinamannersimilartothetransitionofcomputersfromlaboratoriestopersonalhomesandbodies[14].
1.
1HumanCenteredRoboticsOurresearchideologyisinthespiritoftherecentlyidenti-edareaof"humancenteredrobotics"andourapproachtoproblemsoftensharemanythemeswithworkinthiseld.
Ourconjectureisthatbyobservinghumansintheirevery-daylives,awayfrommechanismsandautomation,wecanThisresearchwasnanciallysupportedbyBMDOgrantnum-berN00014-96-1-1200administeredbyONR.
learnvaluableinsightsintothesocialandpsychologicalas-pectsoftheirexistenceandinteractions.
Thesestudieswillinturnmotivatetheformulationofuseful,andhopefullysuccessful,newapplicationsforroboticsresearcherstoad-dress.
Weexpecttodiscovernewapplicationsthathavetra-ditionallyfallenoutsideofwhatisviewedastheroboticseldofstudy.
Inthispaperwemakenoauthoritativeclaimsastothecorrectmethodofthisapproachnordoweproposeageneralsolutiontothisproblem,astherearelikelymany.
Insteadweconcentrateonthedesignofonesuchsystemwhosegoalistoenablepersonaltelepresence.
Ourbeliefisthatfromabriefdiscussionofthehumancentereddesignchoicesthatencompassthisproject,anemerginghumancenteredthemewilldominatethispaper.
1.
2PersonalTele-embodimentOurintentionistoprovidetelepresence1toordinarypeopleinanintuitiveandpersonalmanner.
Inkeepingwithourre-searchparadigm,wefocusnotonthemechanicalelementsofthesystembutonthechoiceandimplementationofspe-cicskillsthatempowerhumanstoexploreandinteractatadistance.
Wedohoweverincludesomediscussionofthemechanicalandroboticcomponentsinthedesign.
Succinctly,weareinterestedinidentifyinganddistillingasmallnumberofhumanbehavioraltraitsorskillsthatareinherenttohumancommunication,understanding,andin-teraction.
Wewillattempttoimplementthesetraitsonin-tuitivehuman-interfaced,networked,mechanicalsystems.
Theultimategoalistoprovideareasonabledegreeofper-sonaltelepresencethatallowshumanstocommunicateandinteractinausefulmannerwithremotepeopleandplacesinwaysbeyondthoseavailablewithcurrentsystems.
Ourclaimisthatsuchsystemscanbebuiltnow,atmini-malcost,andprovidepowerfulnewmetaphorsinmediatedhuman-humancommunication.
Sincethisareahasmanynear-termapplicationsweexpectthatresearcherswillbeabletoexploreawidevarietyoftechniquesforpersonaltele-embodiment.
1Morespecicallywearereferringtopersonaltele-embodiment,tele-robotics,ortele-action.
Thisistoavoidtheam-biguitycausedbythetermtelepresencewhichhasgrowninrecentyearstodescribenotonlysystemsinvolvingdistantrealspaces(i.
e.
tele-robotics)butalsodistantvirtualspacesorVR.
2PreviousandRelatedWorkMethodsofachievingtelepresence2arenotnewwithoneoftherstelectricallycontrolledmechanicalteleoperationalsystemsbeingdevelopedbyGoertz[8]in1954.
Sincethenavarietyofapplicationsfortele-operatedroboticshavebeenexploredbynumerousresearchers.
SpacedoesnotpermitustoincludeasurveyandhencewedeferthereadertoSheri-dan[18].
Mostofthesesystemweredesignedforasin-glespecictaskandarequitecomplex.
Theyalsotypicallyrequirespecialpurposededicatedhardwareandahighlytrainedoperatortocontrolandinteractwiththemechanismintheremoteenvironment.
Inoursystemwestrivedtocon-strainitsdevelopmentsothatitwouldbeaccessibletoawideaudiencewithoutadditional,expensive,orextraordi-naryhardware.
Inessence,telepresenceforthemasses.
TheexponentialgrowthoftheWWWoverthepastsev-eralyearshasresultedinaplethoraofremotecontrolledmechanicaldeviceswhichcanbeaccessedviatheWWW.
Goldberg[9]developeda3DOF(DegreeOfFreedom)teleroboticsystemwhereuserswereabletoexploreare-moteworldwithburiedobjectsand,moreinterestingly,al-teritbyblowingburstsofcompressedairintoitssandlledworld.
Soonafterwards,wedevelopedMechanicalGaze[16],atele-roboticsystemwhereusescouldcontrolacamera'sviewpointandimageresolutiontoobservevari-ousmuseumartifactsplacedwithintherobot'sworkspace.
By1995,Goldberghaddevelopedanotherteleroboticsys-temcalledtheTeleGarden[10]inwhichWWWusersareabletoobserve,plant,andnurturelifewithinalivingremotegarden.
Othershavealsoarguedforahumancenteredapproachtorobotics.
AsAsadapointedoutinhis1997ICRAworkshopentitled"HumanCenteredRobotics,"thereisaoverwhelm-ingneedtodirectroboticsresearchtowardstheneedsofor-dinarypeoplesuchashumanhealthcare,homemedicine,andenhancedcommunicationbetweenpeople.
Similarviewswereexpressedbymanyoftheotherpanelmem-bers[1].
Thisisalsoagrowingbodyofresearchintoubiq-uitoustelepresence[4]andhumancenteredroboticdesignssuchaPeshkin'sCobots[3].
Socialandpsychologicalaspectsofextendedhuman-humaninteractionsmotivatethedesignofourPRoPsandwehaveidentiedawiderangeofresearchinthisarea.
SharedspacesandhumaninteractionwithvideowallssuchastheVideoWhiteboard[20]designedatXeroxPARCandlaterIshii'sClearBoard[11]arefundamentaltodesigningusablePRoPs.
Wearealsointerestedintheuseofvideointele-connectingindividualswhichhasbeennicelyexplored2"Toconveytheideaoftheseremote-controltools,scientistsoftenusethewordsteleoperatorsortelefactors.
Iprefertocallthemtelepresences,anamesuggestedbymyfuturistfriendPatGunkel.
"[15]Figure1:SystemoverviewofatypicalPRoPhardwarecongu-ration.
byKrautandFish[12;7]andothers[6].
WehavealsobeenmotivatedbySteuer's[19]discussionofthedimensionsoftelepresence.
3PRoP:PersonalRovingPresenceAPRoPissimple,inexpensive,internet-controlled,un-tetheredtele-robotthatprovidesthesensationoftele-embodimentinaremoterealspace.
TherstPRoPsweresimplehelium-lledblimpairbornetele-robotscalledspacebrowsers[17].
However,inthispaperwehavechosentofocusonmorerecentlydevelopedterrestrialfour-wheeledPRoPs.
3.
1BasicLayoutTerrestrialPRoPs,sometimesreferredtoassurfacecruis-ersorcarts,aredesignedfromsimple,inexpensiveremote-controlvehicleswithmodicationstoslowthemtohumanwalkingpaceanda1.
5meterverticalpoletoprovideare-alistichumanvantageforthecamera.
On-boardthePRoPisacolorvideocamera,microphone,speaker,colorLCDscreen,afewsimplecustomelectronics,andvariousdriveandservomotors.
ThebasiclayoutforthesystemisshowninFigure1.
Unliketheblimps,thesePRoPscantravelout-doors,requirelessmaintenance,andprovidemuchlongerbatterylife.
TheyalsocarryacompletePCon-boardwithwirelessnetworkinghardwareattached.
Furthermore,weleverageoffofwirelesscommunicationinfrastructuresal-readyinexistence,greatlyextendingtheinhabitableworldofPRoPs.
ArecentlydesignedPRoPisshowninFigure2.
3.
2UserControlAuser,anywhereontheinternet,canuseasimpleJavaappletrunningwithinaJava-enabledbrowsertocontrolthePRoP.
AstheyguidethePRoPforward,backwardsandleft,rightitdelivers,viawirelesscommunications,livevideoandaudiototheremoteoperator'scomputerthroughstandardfreetele-conferencingsoftwarethatrunsonstan-dardpersonalcomputers.
TheremoteoperatorobservestherealworldfromthevantageofthePRoPwhilelisteningtothesoundsandconversationswithincloseproximitytoFigure2:APRoPwithcamerahead,videoLCDscreen,con-trollable"arm/hand"pointer,microphone,speakers,anddrive-ablebase.
it.
Theuserconverseswithgroupsandindividualsbysim-plyspeakingintothemicrophoneconnectedtotheirdesktoporlaptopcomputer,thesounddeliveredviatheinternetandthenawirelesslinktothePRoP'son-boardspeaker.
4HumanCenteredDesignWeshouldstressthatPRoPdesignchoiceswerebasedlargelyonthestudyandobservationofpeopleintheirdailylivesratherthanoninvestigationsintoelaboratehardware.
Itisthismethodologythatwehopetoemphasizeinthispa-perandthusweelaborateonitsroleinthissection.
Initiallyweobservedthatmanyfrequentlyoccurringac-tivitiesinourdailylivesarenotcapturedorconveyedbymoderntechnology.
Despitetheadvancesofourtele-connectedworldoftelephones,pages,cellularphones,andinternetcommunications,wenoticedthatmanysubtle,yetextremelyimportantelementsofhumancommunicationandinteractionsuchasatmosphere,morale,chaos,etc.
weremissingfromtheexperience.
Mostpeoplestillshopbywanderingtheshelves,look-ingforspecials,seeingtheitemtheywant,andaskingaboutitsfeatures.
Wewanderhallwayswithchanceencounterswithpeopleandobjectsplayingasignicantrollinourdailylives.
Oursocialinteractionsarevariegated,andwespon-taneouslymovefromtalkingtooneindividualtoanother,toagroup,toanothergroup,etc.
Inalltheseactivities,oursenses,ourmobility,andoursituatedphysicalformplayes-sentialroles.
However,ourcurrenttechnologicalcommu-nicationchannelsarefartoostructuredtocapturetheseim-portantnuances.
Inthefollowingsubsectionswetracethehumancenteredroboticsmethodologyemployedintheevo-lutionofPRoPs.
4.
1AuralSoundisoneofthemostelementaryandobviousmethodsofhumancommunication.
Therefore,thePRoPdesignin-cludesatwo-way,full-duplexaudiochannelthatallowuserstoengageinremoteconversations.
Oneunexpectedresultofstudyingpeopleusingthisaudiofeaturewastheimpor-tanceofbackground"noise"nearthePRoP.
Theexperi-enceofusingthePRoPwasnoticeablymorecompellingwhenuserswereabletogaugethegeneralmoodofthere-motelocationbyreceivingavarietyofsubtleauralcuessuchasdoorsopening,elevatorsarriving,peopleapproach-ing,nearbyconversations,musicplaying,automobiletraf-c,windblowing,etc.
4.
2VisualDespitethehorricfailureofthePicturephoneofthe1960'sitisclearthatrecentimprovementsinspeed,resolution,andminiaturization,havemadevideoaviableandusefulchan-nelforhumancommunication.
Althoughvideomayaddlit-tleusefulinformationtoatelephoneconversationsbetweenpeople,thevisualappearanceofaremotelocation(color,shape,size,occupancy,lighting,etc.
)isessentialtoconvey-ingseveralofthepreviouslydiscussedintangiblecommuni-cationelementswhentele-visitingaremotelocation.
Againweconsideredtheactivitiesofpeopleandidenti-edtheneedforatleasttwolevelsofvideoresolution.
Thesystemshouldprovideawideangleviewsimilartothehu-maneyefornavigatingandrecognizingpeople(andobjects)andalsoasmallereldofviewforreadingtextonpaper,white-boards,doors,andcomputerscreens.
Wealsonoticedthatwithonlyone-wayvideo,PRoPscouldbemistakenastele-operatedsurveillancetoolsorau-tonomousreconnaissancedrones.
BothofthesetasksarefarfromtheintendedapplicationofPRoPs.
Weremovedthisvideo-asymmetrybyaddingasmall(15cmdiameter)LCDscreenwithavideofeedfromtheremoteuser.
Thistwo-wayvideoisalsoanappropriatemechanismfortransmittingaricherrepresentationoftheremoteuserthroughtheirfacialgesturesandexpressions.
Whenbandwidthisaproblemandthescreenisusedonlytodisplayastillimageoftheremoteuser,wenditstillsucceedsinconveyingtheidentityandexistenceoftheremoteuser.
4.
3MobilityAmbulation,evenwithinasinglebuilding,isasignicantportionofanindividual'sdailyroutineandthusweincludedmobilityasavitalcharacteristicofPRoPs.
Buthowsophis-ticatedshouldthemobilitybeWefoundthatsimplecar-likenavigationofaPRoPonthegroundwasfairlystraight-forwardforausertounderstandandcontrolthougharela-tivelysimpleinterface.
Italsoprovidedenoughfreedomforuserstomaneuverwithin(andoutsideof)buildings.
ThiswasthesimpledesignofourrstPRoP.
However,sincehumaninteractionsoccurwherehumanscantravel,PRoPsmustbeabletoreachmuchoftheworldaccessibletohumans.
Again,wearenotattemptingtocreateanandroidoranthropomorphicrobotsowewillnotattempttohandlewhatwecalldextroushumanmotions.
Inparticu-larweseelittleneedforPRoPstoclimbfences,swingfromropes,leapoverditches,repeldowncliffs,slidedownpoles,etc.
OurbasicphilosophyisthatPRoPsshouldbeabletoac-cessthemajorityoflocationsmosthumansinhabitdaily.
Aimingforsimplicity,wefeelthatPRoPsshouldbeabletoperformsimplelocomotionthroughfairlybenignterrainssuchasmildinclines,curbs,stairs,andsmallvariationsingroundsurface(i.
e.
sidewalks,grass,dirt,etc.
).
Thisin-cludestravelingoutdoorsandalsomeansthatPRoPsmustbebeuntethered(i.
e.
wireless).
Itisalsoimportanttoim-pedetheoverallspeedofthePRoP,typicallythroughvari-ousgearreductions,toroughlymimichumanwalkingpace.
4.
4DirectedGazeWequicklylearnedthatalthoughremoteuserscansee,hear,andmovearound,navigatingstillremainedatedioustaskanddidnotfacilitatetheabilitytoquicklyglancearoundaroomtogetasenseofitssize,occupants,etc.
Thisprob-lemwasremediedbyincorporatingasmallmovable"head"(i.
e.
acameraonacontrollablepan-tiltplatform)ontothePRoP.
OurdeviceissimilartotheGestureCam[13]whichallowsaremoteparticipantinaconversationtohavedirectcontrolofhisorhervisualeldofview.
Thisrelativelysim-plePRoP"head"providesavitallyimportantelementofhumancommunication,directionofattentionorgazeasdis-cussedbyseveralresearchers[5;11].
ThisallowsPRoPstoperformhuman-likeconversationalgesturessuchasturn-ingtofacesomeoneinordertoseethem,addressthem,orjustgiveattentiontothem.
TheseactionsarealsovisibletopeopleinteractinglocallywiththePRoPandprovidesim-plegesturalcuestoletindividualsknowwhentheyarebeingaddressedorlookedatbytheremoteuser.
4.
5PointingandSimpleGesturingBywatchingpeopleinteractwerealizedtheimportanceges-turestoofhumancommunication.
WithourPRoPsremoteusersimmediatelyfoundtheneedtopointoutaperson,ob-ject,ordirectiontotheindividualintheremotespace.
Al-thoughthemovableheadcouldbeusedasacrudesubstitute,itlackedthecorrectvisualgesturalaestheticofpointingandwasoftenambiguoustoindividualswatchingthePRoP.
Weaddedasimple2DOFpointersothatremoteuserscouldpointaswellasmakesimplemotionpatterns.
ThesemotionpatternsallowedthePRoPusertoexpressadditionalnon-verbalcommunicationsgesturessuchasinterestinaconver-sation,agreementwithaspeaker,ortogainattentionforask-ingaquestioninacrowdedroom.
Adequatepointingdoesnotrequireamechanismascom-plexasahumanhand,sinceitisgrossmotionandnotdex-teritythatisneededforthesocialfunctionofgesturing.
Therehasbeenasignicantamountofresearchintoges-turerecognition.
Thesesystemstypicallyaimtoidentifyahumanmotion,typicallymadewithamouse,andinterpretitasaknowngesture.
Forexample,aquickup-downmo-tionofthemousemayberecognizedasthe"scrollpage"gesture.
However,wearemakingaconsciouschoicetousesuchsymbolicdescriptionsofgesturesonlyasalastresort.
InsteadweprefertousecontinuousinputdeviceslikemiceandjoystickstoprovidedirectgesturalinputfromtheusertothePRoP.
Forexample,comparetypingtexttoaspeechsynthesizer,withspokentexttransmittedthroughaspeechcompressionalgorithm.
Thesynthesisapproachmayprovideclean-soundingspeechatlowbandwidth,butallnuanceandemotionalcontentislost.
Similarly,musicwhichisgeneratedbycomputerfromanannotatedmusi-calscoreislifelesscomparedtomusicplayedbyahumanfromthatscore,eveniftherecordingmechanismisidenti-cal(i.
e.
MIDI).
Infactitisnotreallysurprisingthatthroughthesecrudedevicesandnarrowcommunicationchannels,thatrichandcomplexcommunicationispossible.
Recallthatactorstransmittheirgesturestoaudiencememberstensofmetersaway,dancersandmimesworkwithoutspeech,andpup-peteersworkwithoutahumanbodyatall.
Allofususethetelephonewithoutavisualimageofourinterlocutor.
Ourtaskingesturetransmissionistoisolatethekeyaspectsofgesturesoastopreservemeaningascloselyaspossible.
Somefactorsareclearlyimportant,suchastime-stampingtopreservesynchronizationandvelocity.
Others,suchasmap-pinghumandegreesoffreedomtorobot"arm/hand"degreesoffreedomaremuchlessso.
4.
6PhysicalAppearanceandViewpointAlthoughnotanthropomorphic,weobservedthatPRoPde-signislooselycoupledtoafewhuman-liketraitswhichareimportantvisualcuesforsuccessfulcommunicationandin-teraction.
Clearly,asmallground-basedrobotconveysarodent-likeperspectiveoftheworld.
However,alargerobotistypicallyunabletonavigatedownnarrowhallways,passthroughdoors,andimpedesnormalhumantrafcowinabuilding.
Furthermore,largermoreindustrial-typemo-bilerobotsarealsomorelikelytofrightenpeople,detractingfromtheiruseinhumancommunicationandinteraction.
Sincetheystandinasaphysicalproxyforaremoteuser,itmakessensethatPRoPsshouldberoughlythesamesizeasahuman.
Weattacheda1.
5meterverticalpoleatthecen-terofthePRoPtoprovidearealistichumanvantageforthecamera.
IngeneralwehavefoundthatthepositioningofvariousattachmentsonthePRoP(i.
ehead,pointer,arm,etc.
)shouldhavesomecorrespondencetothelocationofanactualhumanbodypartthatprovidestheequivalentfunc-tionality.
Also,allofthecommunicationchannelsshouldbefromthepointofviewofthePRoP(i.
e.
fromon-boardthetele-robot).
ItdoesnotsufcetosimplyhaveacamerasomeplaceintheroomwherethePRoPiscurrentlylocated.
5DiscussionWehavecirculatedourideasonthishumancenteredap-proachtopersonaltele-embodimentandreceivedseveralre-curringcommentsandquestionswhichwewouldliketoad-dress.
PRoPsoundslikejustanotheracronym,wherearethenewideasCertainly,wehesitatetointroduceyetanotherbuzz-wordtotheplethoraoftechno-jargon.
However,itseemsproductivetouseacommontermtodistinguishthegrowingresearchinthisarea.
Obviously,methodsofachievingtelepresencearenotnew,noraresystemsthatallowtele-communication.
Similarly,techniquesandstudiesofhumancommunicationhavebeenexam-inedforcenturies.
Whatwefeelisnewisthemergerofthesemethodsandtheprimaryfocusontheindivid-ualpersontoguidethedesignchoicesoftheentiresys-tem.
Webelievethatevenasmallamountofattentiontothehumanelementinpersonalroboticdesignwillreapcountlessbenets.
Thispaperrepresentsourbestattempttoconveythisdirectionofpersonalroboticsre-search.
Isn'tthisjustanextensionofvideoteleconferenc-ingWhilestandard(andinternet-based)videoteleconfer-encingprovidesanarguablymorerealisticinterfacethanmanyotherformsoftelecommunications,itismoreofanenhancementtoexistingtechnologyratherthananewformofcommunication.
Withvideotele-conferencingwendourselvesxed,staringalmostvoyeuristicallythroughthegazeofanimmovablecam-eraatopsomeone'scomputermonitor.
Asactionsandpeoplepassacrossthecamera'seldofview,wearehelplesstopanandtrackthemorfollowthemintoan-otherroom.
Theresultisa"one-sided"experiencewheretheremoteuserfeelsimmersedbutthereisnophysicalpresenceattheremoteendwithwhichpeo-plecaninteract.
Inessencewestilllackmobilityandautonomy.
Wecannotcontrolwhatweseeorhear.
Evenifwehadcamerasineveryroomandtheabilitytoswitchbetweenthem,theexperiencewouldstilllackthespatialcontinuityofawalkaroundabuilding.
Weclaimthatusersdesireamorerealisticperceptionofphysicalremoteembodiment.
Werealizedtheim-portanceofimmersingthePRoPuserintheremotespacebyprovidingcontinuityofmotionandcontrolofthatmotion.
Theseelementsprovidetheuserthevisualcuesnecessarytostitchtogethertheentirevisualexpe-riencesintoacoherentpictureofabuildinganditsoc-cupantsanddistinguishourworkfromthatofstandardvideoteleconferencing.
Isn'tthisjustanotherformoftelepresenceOurapproachdiffersfundamentallyfrommoretradi-tionalversionsoftelepresencewhichinvolveanan-thropomorphicproxyorandroid.
Instead,PRoPsattempttoachievecertainfundamentalhumanskillswithoutahuman-likeform.
Moreimportantly,ourre-searchisdrivenbythestudyandunderstandingofthesocialandpsychologicalaspectsofextendedhuman–humaninteractionsratherthantheneedtocreateanex-actre-creationoftheremoteexperience.
Forexample,wehavealreadyobservedthatevenwithpoorvideoandcrudemotorcontrols,aPRoPprovidesadequatefunctionalitytoqualifyasausefultoolfortele-visiting.
Whyintroducethetermtele-embodimentPRoPsallowhumanbeingstoprojecttheirpresenceintoarealremotespaceratherthanavirtualspace,usingarobotinsteadofanavatar.
Thisapproachissometimescalled"strongtelepresence"or"tele-embodiment"sincethereisamobilephysicalproxyforthehumanattheendoftheconnection.
Thephysicaltele-robotservesbothasanextensionofitsoperatorandasavisible,mobileentitywithwhichotherpeoplecaninteract.
Wecoinedthetermtele-embodimenttoemphasizetheimportanceofthephysicalmobileman-ifestation.
Idon'twantarobottostandinforme.
Mod-erntechnologyisalreadycreepingintomylifetoomuch.
Wedonotbelievethatwecaneverreplacetruehu-maninteractions,norisitourgoaltodoso.
Insteadweareattemptingtoextendcurrenthumancommuni-cationmethods.
Thatis,ourintentionistoprovidethemeansforindividualstoperformvisitsandinteractionsthatwouldnototherwisebepossibleduetomonetary,time,ordistanceconstraints.
Similarly,itishopedthatvisitsthatnowconsumehoursoftravelingtimecanbetele-conductedinafractionofthetimewithlittlelossofcontent.
Weexpectthistoresultinadditionalfreetimeforindividualstoundertakemorefulllingendeavorsratherthantobeoccupiedsolelywithtrav-eling.
SurebutroboticshasalwaysbeenconcernedwithpeopleTrue,butmuchofthatconcernhasbeendirectedmainlytowardssafetyissueswhenrobotsareoperatingnearhumans.
Insteadweclaimthathumancenteredroboticsfocusesdirectlyonthetasksandissuesthatarepartofdailyhumanactivityrst,beforethedesignoftherobot.
Furthermore,thoseobservationsofpeopledirectlyinuencethedesigndecisionsofthenalsys-tem.
OfcourseweshouldstressthatitisvitalthatsafetybeaprimaryconcernwhendesigningPRoPs.
Wepro-poseateleoperationalvariationonAsimov'srstlawofrobotics3whichstipulatesthatatnotimeshouldaPRoPeverbecapableofinjuringahumanbeing,re-gardlessoftheactionorinactionoftheremotetele-operator.
6ConclusionOurclaimisthatPRoPsprovideanextremelyuseful,func-tional,powerfulnewtoolforsupportinghumancommuni-cationandinteractionatadistance.
Theyenableavarietyofimportantworkandsocialtele-activitiesfarbeyondwhatweperformcurrentlywithourcomputersandnetworks.
PRoPsarealsoanidealplatformforstudyingcomputer-mediatedhumaninteractionbecausetheyoperateinexist-ingsocialspacesandcaninteractwithgroupsofhumans.
DespiteourlimitedexperienceusingPRoPs,wehavebeenabletoidentifyseveralfactorsthatweconsidervitaltopro-vidingthemostcompellingoverallexperienceforboththeremoteandlocalusers.
Thisiswhyourresearchdrawsasmuchonthesociologyofgroupinteractionsasonsensingandactuationtechniques.
Infactweneedtheformertodriveourchoicesforthelatter.
Furthermorewebelievethatroboticsasaresearchareaispoisedtobeginsignicantcontributionsintothedailylivesofpeopleandsocietyinwaysthatwearelikelytonotyetevenimagine.
Welikenthistothemovementoftheelabo-rateinstitutionalcalculatingenginesofonlyafewdecadesagointothecasualdailyinteractionsweobservebetweenhumansandcomputerstoday.
3"Arobotmaynotinjureahumanbeingor,throughinaction,allowahumanbeingtocometoharm.
"HandbookofRobotics,56thEdition,2058A.
D.
,asquotedinI,RobotbyAsimov[2]Weclaimthatpersonaltele-embodimentisanexampleofhumancenteredrobotics.
Mostimportantly,weemphasizeanddemonstratetheimportanceofconductingroboticsre-searchthatfocusesontheindividualpersonratherthantheintricatecomplexitiesofthemachineandcallforresearchfollowingthismethodology.
References[1]ASADA,H.
H.
,BEKEY,G.
,TARN,T.
J.
,KAZEROONI,H.
,LEIFER,L.
,PI-CARD,R.
,ANDDARIO,P.
Workshop:Humancenteredrobotics.
InIEEEInternationalConferenceonRoboticsandAutomation(April1997).
[2]ASIMOV,I.
I,Robot.
GardenCity,N.
Y.
:Doubleday,1950.
[3]COLGATE,J.
E.
,WANNASUPHOPRASIT,W.
,ANDPESHKIN,M.
A.
Cobots:Robotsforcollaborationwithhumanoperators.
InIMECE(1996).
[4]DOHERTY,M.
,GREENE,M.
,KEATON,D.
,OCH,C.
,SEIDL,M.
,WAITE,W.
,ANDZORN,B.
Programmableubiquitousteleroboticdevices.
InPro-ceedingsofSPIETelemanipulatorandTelepresenceTechnologiesIV(October1997),vol.
3206,pp.
150–157.
[5]DONATH,J.
Theillustratedconversation.
InMultimediaToolsandApplications(March1995),vol.
1.
[6]FINN,K.
,SELLEN,A.
,ANDWILBUR,S.
,Eds.
Video-MediatedCommunica-tion.
L.
ErlbaumAssociates,1997.
[7]FISH,R.
,KRAUT,R.
,ROOT,R.
,ANDRICE,R.
Evaluatingvideoasatech-nologyforinformalcommunication.
CommunicationsoftheACM36,1(1993),48–61.
[8]GOERTZ,R.
,ANDTHOMPSON,R.
Electronicallycontrolledmanipulator.
Nu-cleonics(1954).
[9]GOLDBERG,K.
,MASCHA,M.
,GENTNER,S.
,ROTHENBERG,N.
,SUTTER,C.
,ANDWIEGLEY,J.
Robotteleoperationviawww.
InInternationalConfer-enceonRoboticsandAutomation(May1995),IEEE.
[10]GOLDBERG,K.
,SANTARROMANA,J.
,BEKEY,G.
,GENTNER,S.
,MORRIS,R.
,SUTTER,C.
,WIEGLEY,J.
,ANDBERGER,E.
Thetelegarden.
InSIG-GRAPH(1995).
[11]ISHII,H.
,ANDKOBAYASHI,M.
Clearboard:Aseamlessmediumforshareddrawingandconversationwitheyecontact.
InACMSIGCHI(1992).
[12]KRAUT,R.
,FISH,R.
,ROOT,R.
,ANDCHALFONTE,B.
Informalcommuni-cationinorganizations:Form,function,andtechnology.
InACMConferenceonComputer-SupportedCooperativeWork(1988).
[13]KUZUOKA,H.
,KOSUGE,T.
,ANDTANAKA,M.
Gesturecam:avideocom-municationsystemforsympatheticremotecollaboration.
InACMConferenceonComputer-SupportedCooperativeWork(October1994).
[14]MANN,S.
Smartclothing:Thewearablecomputerandwearcam.
PersonalTechnologies(March1997).
Volume1,Issue1.
[15]MINSKY,M.
Telepresence.
Omni2,9(June1980).
[16]PAULOS,E.
,ANDCANNY,J.
Deliveringrealrealitytotheworldwidewebviatelerobotics.
InIEEEInternationalConferenceonRoboticsandAutomation(1996).
[17]PAULOS,E.
,ANDCANNY,J.
Ubiquitoustele-embodiment:Applicationsandimplications.
InternationalJournalofHuman-ComputerStudies/KnowledgeAcquisition(1997).
SpecialIssueonInnovativeApplicationsoftheWorldWideWeb.
[18]SHERIDAN,T.
B.
Telerobotics,AutomationandHumanSupervisoryControl.
MITPress,1992.
[19]STEUER,J.
Deningvirtualreality:Dimensionsdeterminingtelepresence.
JournalofCommunication4(Autumn1992).
[20]TANG,J.
,ANDMINEMAN,S.
Videowhiteboard:Videoshadowstosupportremotecolaboration.
InACMSIGCHI(1991).