AAWA paper master v12

Ship Intelligence

Remote and Autonomous Ships The next steps

RemoteandAutonomousShip–Thenextsteps

AAWA Position Paper © Rolls-Royce plc Registered office: 62 Buckingham Gate, London, SW1E 6AT. Company number 1004142. Registered in England

2

TABLEOFCONTENTS Page Author(s)

Introduction

Background

AAWAInitiative

Visionofremotecontrolledshipoperation

Voyageplanningandinitiation

Unmooringandmanoeuvringoutofharbour

Operationmodesatopensea

Portapproachanddocking

Applicabilityfordifferentshiptypes

Conclusion

3

4

5

6

8

9

10

12

12

13

EsaJokioinen

Technologiesformarinesituationalawarenessandautonomousnavigation

Autonomousnavigationofthevessel

Situationalawareness(SA)forautonomousships

Off-shipcommunication

15

18

23

30

JonnePoikonen

MikaHyvönen

AnttiKolu

TeroJokela

JariTissari

AriPaasio

LegalImplicationsofremoteandautonomousshipping

Introduction

Lawatsea

Technicalrequirements

Liabilityrules

Summary

35

36

36

40

49

54

HenrikRingbom

FelixCollin

MikaViljanen

Safetyandsecurityinautonomousshipping–challengesforresearchand

development

Introducingofautonomousmerchantshipsformaritimeoperation

Are‘unmannedships’safe?

Preconditionsofsafetyandsecurity

Focalareasofrisk–someselectedexamples

Managingshippingsafetyandsecurityinshortandlongterm

Buildingriskunderstandingforthefuture

Recommendations

56

58

59

60

62

68

72

72

RistoJalonen

RistoTuominen

MikaelWahlström

FromInnovationstoMarkets–RedefiningShipping

Redefiningshipping–atransitiontoautonomousshipping

Autonomousshipping–anissueofbusinessrelationshipsand

networks

Autonomousshipping–arenewedsetofrolesbetweenthekey

actors

Transitiondriverstoautonomousshipping

Transitionroadmap

Conclusion

74

75

77

79

80

82

85

JouniSaarni

SiniNordberg-Davies

HannuMakkonen



3

IntroductionEsaJokioinen–Rolls-Royce–HeadofBlueOceanTeam



4

1.Background

“Autonomousshippingisthefutureofthemaritimeindustry.Asdisruptiveasthesmartphone,the

smartshipwillrevolutionisethelandscapeofshipdesignandoperations”

MikaelMakinen,PresidentRolls-RoyceMarine

Tenyearsagotheveryideathatyoucouldmanageyourlifethroughasmallglassscreen,was

consideredalmostimpossible.Nowfewofuswouldwanttobewithoutone.Twoyearsagotalkof

intelligentshipswasconsideredbymanyasafuturisticfantasy.Today,theprospectofaremote

controlledshipincommercialusebytheendofthedecadeisareality.

Thetechnologies,particularlysensortechnologies,neededtomakeremoteandautonomousshipsa

realityalreadyexist.Thechallengeistofindtheoptimumwaytocombinethemreliablyandcost

effectively.Thedecisionalgorithmswhichwillhelpsuchvesselsdecidewhatactiontotakeinthelight

ofthatsensorinformationarebeingperfected.Thisrequiresaninterpretationofmaritimerulesand

regulationsleadingtochallengesofinterpretationfortheprogrammer.Thedevelopmentofdecision

supportsystemswillbeagradualanditerativeprocesssubjecttoextensivetestingandsimulation.

Tosecureregulatoryapproval;aswellasindustrysupportandpublicacceptance,remoteand

autonomousshipswillneedtobeasleastassafeasexistingvessels.Theyhavethepotentialtoreduce

humanbasederrorsbutatthesametimenewtypesofriskwillariseandwillneedtobeaddressed.A

comprehensiveandstructuredwaytoidentifyandaddresstheserisksisrequired.

Unmannedshipsopenupexcitingpossibilitiestoredefinethewayashipisdesignedandfunctions.

Whentherearenopeopleonboard,manyconstraintsontheshiplayoutareremoved.Oneofthemost

obviousistheremovaloftheaccommodationandwiththattheentiredeckhouse.Thiswillsavecost,

weightandspace,aswellasenablingtheshiptocarrymorecargo.Ashipcontainssystemsthatare

onlytheretoservethecrew.Theirremovalwillsimplifytheentireship,whichshouldimprove

reliabilityandproductivitywhilereducingbuildandoperatingcosts.

Futurevesselswillstillneedhumaninputfromlandmakingconnectivitybetweentheshipandthe

shorecrucial.Suchcommunicationwillneedtobebidirectional,accurate,scalableandsupportedby

multiplesystemscreatingredundancyandminimisingrisk.Sufficientcommunicationlinkcapacityfor

shipsensormonitoringandremotecontrol,whennecessary,hastobeguaranteed.Continuous,

guaranteedconnectivitygivesustheabilitytomonitorequipmentinserviceinrealtimedetecting,



5

diagnosingandprioritisingissueswithcriticalequipmenthelpingcustomersgetthemostoutoftheir

assetsbyoptimisingbothoperationsandmaintenanceschedules.

Sucharichstreamofdataandmorestandardisedshipswillhaveenormousconsequencesforthe

shippingindustry.

Itwillallowshipownerstomanagetheirfleettooptimiseoperationsandmaximiseprofit.Bylooking

atdatafromindividualshipstogethertheywillbeabletoidentifythebestcombinationofroute,cargo,

maintenancescheduleandfuelpriceforthefleetasawholegettingthemaximumvaluefromasetof

veryexpensiveassets.

Inthisshipownerswillnotbealone.Increaseddigitalisationwillcreatenewshippingservices,suchas

moreefficientpoolingandalliances,leasingofassets,onlinecargoservicemarketplaces,etc.Someof

theseserviceswillsupportexistingmarketplayersandsomewillbedisruptive–allowinganew

playertoenterthemarketandtakeoverlargesharesofthebusinessinthesamewayasUber,Spotify

andAirbnbhavedoneinotherindustrysectors.

Rolls-RoycetogetherwiththeotherpartnersintheAAWAproject,DNVGL,Inmarsat,Deltamarin,

NAPA,BrighthouseIntelligence,FinferriesandESLShipping–andwiththesupportofTekesRolls-

Royce–isleadingthisrevolution.

2.AAWAInitiative

TheAdvancedAutonomousWaterborneApplications(AAWA)Initiativeisa€6.6millionproject

fundedbyTekes(FinnishFundingAgencyforTechnologyandInnovation)aimstoproducethe

specificationandpreliminarydesignsforthenextgenerationofadvancedshipsolutions.

Itbringstogetheruniversities,shipdesigners,equipmentmanufacturers,andclassificationsocietiesto

exploretheeconomic,social,legal,regulatoryandtechnologicalfactors,whichneedtobeaddressedto

makeautonomousshipsareality.

Theprojectwillrununtiltheendof2017andwillpavethewayforsolutions-designedtovalidatethe

project’sresearch.TheprojectwillcombinetheexpertiseofsomeofFinland’stopacademic

researchersfromTampereUniversityofTechnology;VTTTechnicalResearchCentreofFinlandLtd;

ÅboAkademiUniversity;AaltoUniversity;theUniversityofTurku;andleadingmembersofthe



6

maritimeclusterincludingRolls-Royce,DNVGL,Inmarsat,Deltamarin,NAPA,Brighthouse

Intelligence,FinferriesandESLShipping.

Thewiderangingprojectlooksatresearchcarriedouttodatebeforeexploringthebusinesscasefor

autonomousapplications,thesafetyandsecurityimplicationsofdesigningandoperatingremotely

operatedships,thelegalandregulatoryimplicationsandtheexistenceandreadinessofasupplier

networkabletodelivercommerciallyapplicableproductsintheshorttomediumterm.The

technologicalworkstream,ledbyRolls-Royce,encompassestheimplicationsofremotecontroland

autonomyofshipsforpropulsion,deckmachineryandautomationandcontrol,using,wherepossible,

establishedtechnologyforrapidcommercialisation.

Forremotecontrolledandautonomousshipstobecomearealityanumberofcriticalquestionsneed

tobeanswered:

• Whattechnologyisneededandhowcanitbebestcombinedtoallowavesseltooperate

autonomouslyandmilesfromshore;

• Howcananautonomousvesselbemadeatleastassafeasexistingships,whatnewriskswillit

faceandhowcantheybemitigated;

• Whatwillbetheincentiveforshipownersandoperatorstoinvestinautonomousvesselsand

• Areautonomousshipslegalandwhoisliableintheeventofanaccident?

In2015thefirstphaseoftheprojecthasexaminedthecurrentstateofthemaritimeindustryand

whatcanbelearntfromotherindustries–fromaviation’sdronesanddriverlesscarstothe

smartphone.Theprojecthasexploredthecurrentstateofunderstandingofthetechnological,safety,

legalandeconomicaspectsofremoteandautonomousoperation.Thefindingsofthisresearchcanbe

foundinthiswhitepaper.

ThenexttwophasesofAAWAwillbuildonthefindingsfromthefirstphasetodevelopthetechnical,

legalandsafetyspecificationsforaproofofconceptdemonstratorbytheend

of2017.

3.VisionofremotecontrolledshipoperationTheconceptofdynamicautonomy

Therearenumberofdifferentdefinitionsofautonomyandmachineintelligenceintheliterature.

Levelsofautonomy(LOA)areoftenusedtodescribetowhatdegreethemachinecanactonitsown.

Probablythemostwell-knowndescriptionsforLOAaredevelopedbyThomasSheridan.TheSheridan



7

scaleincludesacontinuousrangeofdefinitionsfromamachinebeingcompletelycontrolledbyhuman

(i.e.tele-operated)throughthemachinebeingfullyautonomousandnotrequiringanyinputfromthe

humanbeforetakingactions.

Table1–Sheridanlevelsofautonomy

Differentvariationsofthistypesofscalehavebeendevelopedintheresearch.Acommonconclusionis

thatsuchscalesmaynotbeapplicabletoentireoperationbutaremostusefulwhenappliedto

differentsubtasksoftheautonomousmachine.

Thisconclusionisalsohighlyrelevantforautonomousshipsasthebehaviourofthevessel(i.e.LOA)

andrequiredamountofhumaninteractionwilldependentonthestateofthevesselandsubtaskbeing

executed.Thistypeof“adjustable”or“dynamic”autonomyisaconceptwhichisoftendiscussedin

contextofmobilerobotsinwhichthemachinecanbeoperatedforperiodsoftimeonitsown

dependingonthelimitsgivenforthedecisionmakingtolerance.Therobotcanhandlesimpletasks

autonomouslybutwhenthetasksaregettingmorecomplexincreasinginteractionwiththehuman

operatorisneeded.

Remotecontrolledshipswillfollowthistypeofdynamicautonomyapproachdependingonthestateof

thevesselandmissionbeingexecuted.Insomecases,suchasnavigationintheopenseas,theshipcan

benearlyfullyautonomouswhereasforsomepartsofthevoyageitwillrequireclosesupervisionand

decisionmaking,orevenfulltele-operationfromthehumanoperator.

Inordertounderstandhowautonomousshipswouldwork,anexampleofgeneralcargovessel

operatingbetweentwoportsisdescribedinthefollowing.Theexamplewillshowexamplesof

differentlevelsofautonomyduringdifferentphasesofthevoyage.Forsimplicityasinglevessel

operatedbyonehumanoperatorispresented.

Level Description10 The computer does everything autonomously, ignores human

9 The computer informes human only if it (the computer) decides so8 The computer informes human only if asked7 The computer executes automatically, when neccssary informing human6 The computer allows human a restricted time to veto before automatic execution5 The computer executes the suggested action if human approves4 Computer suggests single alternative3 Computer narrows aleternatives down to a few2 The computer offers a complete set of decision alternatives1 The computer offers no assistance, human in charge of all decisions and actions



8

4.Voyageplanningandinitiation

Therearecertainthingswhicharerelatedtoremotecontroloperationthathavetobetakeninto

accountbytheoperatorwhileplanningthevoyageormissionforthevessel.Autonomousvesselswill

useamixofdifferentsatelliteandlandbasedcommunicationnetworksdependingontheir

availability,qualityandprice.Highbandwidthsatellitecommunicationsystemsprovidethecapability

tooperateanautonomousvesseldespitethelocationinvastmajorityofautonomousoperation

modes.Howeversomeoftheremotecontrolorremotesupervisionmodesmightrequirealatencyand

bandwidththatexceedsthecapabilityofthesatellitesystemsinadverseweatherconditions.The

operatorwillhavetoensurethatthereissufficientconnectivityfortheintendedmission.Evenifdata

transferofautonomousshipshashighestpriorityinthesenetworkstheoperatorwillhavetoreview

thetrafficandweatherconditionsinordertodecidewhatistheprimaryoperationstrategyforeach

leg.

Fromvoyageplanningpointofviewthismeansdefiningwhichlegsshallbeoperatedinremote

controlandwhichareexecutedautonomously.Oncethisdecisionhasbeenmade,theoperatorwill

havetofurtherdefinenavigationalstrategiesalongwithfallbackstrategiesforeachleg.Thefallback

strategysequenceisexecutedonlyiftheshipexperiencesanunexpectedreductioninconnectivity

simultaneouslywithoperationalchallengewhichwouldnormallyrequireoperatorintervention.

Thefallbackstrategycouldinclude:askingoperatortotakemanualcontrol(iffailed),slowdownand

proceedtofollowingwaypoint(iffailed),stopthevesselandstayinDPmode(iffailed),navigateto

previouswaypoint(iffailed),navigatebacktopresetsafelocation.Thecommandsandtheirexecution

sequenceisobviouslynotsameinallpartsofthevoyage.Forexampletryingtomaintainitspositionin

themiddleofacongestedandnarrowfairwayinharshweathermightnotbeafeasiblestrategy.The

voyageplanaswellasthefallbackstrategiescanalwaysbemodifiedduringthevoyageusingthe

satellitecommunicationlink.

Theshipwillalsoneedtohaveanautomaticsystemforverifyingtheseareadinessbeforestartingthe

voyage.Mostofthesystemscanbecheckedremotelybytheoperatorwhileinsomeareas(suchas

securingcargo)shorebasedcrewcanalsobeusedtocheckthatvoyagecanbestarted.

5.Unmooringandmanoeuvringoutofharbour

Themooringsystemsforanautonomousvesselcanbefullyorsemi-automatic.Inthecaseofafully

automaticmooringsystemthecompletemooringandunmooringoperationcanberemotecontrolled



9

orisautomaticallyexecutedbytheautonomousvessel.Semi-automaticmooringmeansthat

connectiontothequaycanbemadeautomaticallybutcrewisneededtosecurethedocking(i.e.using

conventionalrope-basedsystems).Bothoftheserequirepotentiallysomemodificationstothe

docksideinfrastructurewhichmeansthattheeconomicfeasibilityofthemooringsystemwilldepend

onhowmanyvesselsareabletousethesamedockingsystem.Solutionsforthisexistinthemarket

andAAWAexplorestheirfeasibilityforautonomousvesselsalongwithdevelopmentofnewpotential

automaticmooringarrangements.

Figure1–Semi-automatedmooringsystem

Whentheshipismanoeuvredoutofthecongestedharbourareatheoperatorcaneitherhavedirect

remotecontrolorsupervisorycontrolwhichissupportedbytheonboardsituationawareness

systems.Inthistypeofoperationahighbandwidthandlowlatencycommunicationlinkisneeded.In

certainareasthiscanbeprovidedbytheland-basedcommunicationnetworksandsatellite

communicationsystemsremainasback-up.

Figure2–Supervisoryteleoperation



10

Remotecontrolcanmeandirectjoystick-typeoperationmodesalreadyexistinginthedynamic

positioningsolutionssuchaslockingspeed,headingorrelativepositiontoanobjectareavailable.

However,amorepracticalwayincaseofmostshiptypesiscontrollingthevesselbysendingwaypoint

andthedynamicpositioningcontrolcomputerandautonomouscontrolsystemtakescareofactual

propulsioncontrols.Insomeareasitispotentiallyalsofeasibletogodirectlytoautonomousmode

insteadofstartingwithteleoperationorsupervisorycontrol.

6.Operationmodesatopensea

Innormalautonomousmodetheshipexecutestheplannedmission(e.g.navigationtothenext

waypoint)accordingtothedefinedplan.Inthismodethedatatransferbetweentheshipandoperator

isminimisedandlimitedtoonlyrelevantstatusdatasuchasship’slocation,heading,speed,ETAto

nextwaypoint(orareaofclosersupervision)andkeyinformationfromthesituationalawareness

systemsaswellascriticalshipsystems.Whiletheinteractionrequirementbetweentheshipand

operatorisminimalinthisnormalstate,itispossiblefortheoperatortosupervisemorethanone

vesselatthetime.Thismeansthattheautonomylevelofthevesselishighaslongasthemission

executionisproceedingaccordingtotheplanmadebytheoperator.

Additionalinformationwillbeprovidedautomaticallyincasethesituationalawarenesssystemsand

theautonomousnavigationsystemautonomousdecisionmakingthresholdisexceededanduser

notification,confirmationorinterventionisrequired.Thismeansthattheautonomylevelis

dynamicallyadjustedifthemissionexecutionisnotproceedingaccordingtotheoriginalplanandthe

autonomousnavigationsystemseesthatadjustmentsareneeded.

Differentlevelsofoperatorinteractionwillberequesteddependingontheoperationalscenario.For

exampleifthevesselisdeviatingfromtheplannedcoursebetweenthetwowaypointsbutstayswithin

specifiedmarginstheautonomousnavigationsystemonlynotifiestheoperatoraboutplannedevasion

andgivestheoperatorapossibilitytovetoforalimitedtime.Oneexampleofsuchevasioncouldbe

takingautomaticactiontokeepoutofthewayofanothervesselbyslightlychangingtheheadingor

speed.TheoperatorcouldchoosetouseVHFradiotocommunicatewiththeothervesselandconfirm

thatactiontakenbythevesselissafeforbothparties,andifmodificationsareneededtheoperatorcan

takethevesselinmanualcontrol.

Amorecomplicatedcaserequiringuserdecisionmakingiswhenthevesselneedstochangethe

courseinsuchawaythatcompletewaypointhastobere-planned(e.g.evasionoroffsetfromthe

plannedpathisnotenoughtosolvethenavigationalchallenge).Inordertoensurethatchangestothe



11

planaremadeinasafewayoperatorconfirmationwillberequested.Theautonomousnavigation

systemwillofferoneormorealternativesofhowthewaypointcouldbemodifiedbuttheoperatorwill

finallymakethedecisionhowtocontinuethevoyage.

Itcanalsobeexpectedthattherewillbecomplexscenarioswheretheautonomousnavigationsystem

pathplanningandalgorithmscannotunambiguouslysolvethesituation.Exampleofthiscouldbeif

extremelylargenumberofcraftsorotherobjectsaredetectedandthepathplanningalgorithmsare

notcapabletoidentifythemandtherebythesystemcannotdeterminehowthenavigationshould

proceed.Inthistypeofscenariothevesselwillimmediatelysenda“pan-pan”messagetotheoperator

indicatingthatitisinurgentneedofassistance.Theshiphaspredefinedsetoffallbackstrategiesthat

itwillstarttoexecuteintheplannedorderifuserresponseisnotreceived,andin“pan-pan”-

dependingontheurgency,automaticfallbackstrategyexecutioncanalsobestartedimmediately.

Figure3–Differentscenariosrequiredifferentlevelsofoperatorinvolvement

Operationoftheautonomousvesselwillcombinedifferentautonomylevelsdynamicallydependingon

thestateofthevesselandexternalconditions.Obviouslyasthecontrolalgorithmswillevolveand

matureovertime,theshipswillbecapableofhandlingincreasinglycomplexsituationsontheirown.

Whentheautonomousshipfleetincreasesitwillalsobepossiblethattheautonomousshipsshare

voyageplansandcommunicatewitheachotherautomaticallywhichreducestheoperatorload.

However,therewillalwaysbemannedvesselssailingalongwithautonomousshipswhichmeansthat

humanoperatorwillbenecessaryforquitesometimetointerpretthisinformationuntilclear

standardsforinformationsharingbetweenmannedandunmannedvesselsaredeveloped.



12

7.Portapproachanddocking

Whenapproachingtheportareatheoperatorcanagainchoosetotaketeleoperationtypecontrolor

increasethesupervisionlevelofthevessel.ThismightbenecessaryfromVTSpointofview,butalso

becausepilotingmightberequired.

Pilotingcaninthefuturebeorganisedinnumberofdifferentwaysforautonomousvessels.One

alternativeisthatthepilothascapabilitiestotakecontroloftheautonomousvessel,oralternatively

theautonomousvesseloperatorcanholdapilotlicensefortheintendedoperationareas.

Implementationofautonomousvesselswillmostlikelystartfromnationalorregionalwatersand

frequentrouteswhichmeansthatpilotingproceduresandpracticalitieswithVTScanbeagreedcase-

by-caseforthefirstvessels.

Whenoperatingthevesselinproximityoftheshoreitisagainpossibletorelyontheland-based

systemsforcommunication.Additionallythenavigationsystemcanuselandbasedexternalreference

systemsforpositioningwhichwillbeusefulespeciallyinportareas.Inadditionlandbasedcamera

andradarsystemscanbeusedtonavigatethevesselsafelyalongsidethedock.

8.Applicabilityfordifferentshiptypes

Theexampledescribedintheearlierchaptergivesanideaofhowdynamicautonomywouldworkfor

shipoperations.Obviouslytypeandlevelofautonomywillbealsohighlydependentontheshiptype,

size,operationalareaandconditions.Forexampleanautonomoustugwouldfollowthesame

principlesbutastheoperationismuchmorefocusedaroundthetowingmission,thecontroland

autonomyprincipleshavetobedefinedfromadifferentpointofview.

Generallyspeakingthemorevariationsandcomplexitythemissionhas,themoretheshipwillhaveto

relyonoperatorassistanceandremotecontrolatleastinthefirstphasesoftheimplementation.

Anotherexamplecouldbeaninlandferrymakingtensofidenticalcrossingseveryday.Inthiscasethe

missioninitselfhasmuchlessvariationandtheautonomylevelinexecutingthetaskcanbemuch

higher.Atthesametimeitisimportanttokeepinmindthateventhoughthebasicmissionisnot

variedtoomuch,theconditionssuchasweatherandtrafficcanchangeconsiderably.Onboardcrew

mightstillbeneededinthesecasestosupervisesafetyoftheoperationseveniftheshipexecutesthe

basicmissionnearlyautonomously.



13

Inadditiontodifferencesinoperationandconditions,therearealsobigdifferencesinhowtheships

willreacttocontrolcommands.Alargecontainervesselandsmallgeneralcargovesselwillneedto

havetheirownship-specificmodelsofcontrolalgorithmseventhoughthefundamentalsofhowthey

reactautonomouslytodifferentnavigationalconditionswouldbefollowingthesameprinciples.

Technicallythisalsomeansthatthesituationalawarenesssystemwillhavetobedifferentasthe

reactiondistance(time)ofalargevesselisconsiderablyhigherandhigherpredictabilitylevelsare

needed.

9.Conclusion

ThefirstphaseoftheAAWAprojecthasexaminedthecurrentstateofthemaritimeindustryandwhat

canbelearntfromotherindustries.Theprojecthasexploredthecurrentstateofunderstandingofthe

technological,safety,legalandeconomicaspectsofremoteandautonomousoperation.

Theinitialconclusionsare:

1. Therewillbenosingleremoteorautonomousshipsolutionbutratherahybridofthetwo

whichwilldependonthetypeandfunctionofthevessel.

2. Thetechnologiesneededtomakeremoteandautonomousshipsarealityexist.Thechallenge

istofindtheoptimumwaytocombinethemreliablyandcosteffectively.Thedevelopmentof

decisionsupportsystemsforautonomousvesselswillbeagradualanditerativeprocessand

subjecttoextensivetestingandsimulation.

3. Theoperationofremoteandautonomousshipswillbeasleastassafeasexistingvessels.

Thereispotentialtoreducehumanbasederrorsbutatthesametimenewtypesofriskwill

ariseandwillneedtobeidentifiedandaddressed.

4. Legislationcanbechangedifthereisapoliticalwill.Forremoteandautonomousshippingto

becomearealityeffortisneededatallregulatorylevels.Thelegalchallengesofconstructing

andoperatingademonstrationvesselatanationallevelneedtobeexploredwhilst

simultaneouslyconsideringappropriaterulechangesattheIMO.Questionsofliabilityfor

autonomousshipsaresubjecttonationalvariations,butgenerallyitseemsthatthereisless

needforregulatorychangeinthisfield.Whatneedstobeexplored,however,istowhatextent

otherliabilityrules,suchasproductliability,wouldaffecttraditionalrulesofmaritimeliability

andinsurance.

5. Remoteandautonomousshipshavethepotentialtoredefinethemaritimeindustryandthe

rolesofplayersinitwithimplicationsforshippingcompanies,shipbuilders,maritimesystems



14

providersandtechnologycompaniesfromother(especiallytheautomotive)sectors.

Thenextstepsare:

• thedevelopmentandtestingofspecifictechnologicalsolutionsforautonomousoperations

usingbothsimulatorsaswellastestsatseaacrossavarietyofenvironmentalconditions-the

optimumwaytocombinethedifferentsensortechnologiesinarangeofoperatingandclimatic

conditionswillbethesubjectofaseriesofteststhisyearonboardtheFinFerriesvessel,the

Stella,operatingbetweenKorpoandHoutskär;

• researchtounderstandthechangedandnewrisks(avarietyofknownandunknownhazards)

presentedbynewandemergingtechnology,buildingonthemarineindustry’sexperienceof

systematicandcomprehensiveriskassessments,todevelopnewapproaches;

• exploringthelegalchallengesofconstructingandoperatingademonstrationvesselata

nationallevelwhilstsimultaneouslyconsideringappropriaterulechangesattheIMO;

• exploringstakeholderviewsofremoteandautonomousshippingtoestablishcostandrevenue

modelsofautonomousoperationfordifferentshiptypes.

TheoutcomePhaseIIwillbethetechnical,legalandsafetyspecificationsforafullscaleproofof

conceptdemonstratorbytheendof2017andaremotecontrolledshipincommercialusebytheend

ofthedecade.

Therevolutionhasbegun.



15

TechnologyJonnePoikonen,SeniorResearchFellow,D.Sc.(Tech.),UniversityofTurku

MikaHyvönen,SeniorResearchFellow,D.Sc.(Tech.),TampereUniversityof

Technology

AnttiKolu,DoctoralStudent,M.Sc.(Tech.),TampereUniversityofTechnology

TeroJokela,SeniorResearchFellow,D.Sc.(Tech.),UniversityofTurku

JariTissari,ProjectResearcher,M.Sc.(Tech.),UniversityofTurku

AriPaasio,Professor,D.Sc.(Tech.),UniversityofTurku



16

Technologiesformarinesituationalawarenessandautonomous

navigation

Technologiesforrealisingremoteandautonomousshipsexist.Thetaskistofindtheoptimumwayto

combinethemreliablyandcosteffectively.

Thedevelopmentofautonomousvehicles,eitheronland,airorseahasseengreatprogressduringthe

last10years.Thishasbeenenabledbyadvancesintechnologies,whichenableperceptionofthe

surroundingenvironment,pathplanningandvehiclecontrolinrealtime.Withacombinationofan

arrayofadvancedsensortechnologies–becomingavailablealsobeyondearliermilitaryandscientific

use–andrapidlyincreasingdataprocessingperformance,wehavereachedatechnologicallevelon

whichfullvehicularautonomyisindeedfeasible.

Themostprogresshasbeendemonstratedinthefieldofautonomouscars.Thisisnaturalduetothe

largemass-marketpotentialandtheglobalneedforincreasedtrafficsafety.Formilitaryapplications,

significantresearchanddevelopmenthasbeencarriedoutinthefieldsofautonomouslandvehicles,

aviationandalsomarinevessels,suchassmallpatrolandattackboats[Elkins,2010].Recently,efforts

tocreatesolutionsalsoforcivilianautonomousmarinevesselshaveseenasignificantincrease,e.g.in

theformofmanynewresearchprogramsinco-operationwithacademiaandmarinetechnology

companies,includingAAWA.

Oneofthekeytechnologiesforanyreliableautonomousvehiclenavigationissensorfusion.When

creatingSituationalAwareness(SA)foranautonomousvehicle,nosinglesensortechnologycan

providesufficientperformanceunderallpossibleconditions.Therefore,inordertoguaranteethatthe

informationonthevehicle’ssurroundingsissufficientlyaccurateatalltimes,theinputfrommultiple

sensorshastobecombinedandanalysed.Thesensordataprocessingshouldthenbeseamlessly

integratedwithsubsequentpathplanningandreactivecollisionavoidancesystems,whichmaintaina

constantlyupdateddetailedmapofthevehicle’senvironment,allowingthevehicletoplanitsroute

andavoidanycollisionswithobjectsorothervehicles.Themapgatheredfromsensordatacanalsobe

augmentedwithdatafromstaticmapdatabasessuchasGoogleMapsforcarsorelectronicnautical

charts(ECDIS)forships,whichpresentstaticobjectsofthesurroundingarea.

Researchonautonomouscarsoffersthemostextensivesourceofpubliclyavailableinformationon

technologiesdevelopedforautonomousvehicles.ThetoppartofFigure1illustratesatypicalsensing



17

andprocessingpipelineappliedinautonomouscars.Multiplesensorsareusedtoextractdatafrom

thesurroundingsofthevehicle.Thepreferenceondifferentsensortypesvaries:GoogleappliesLIDAR

asthemainsensorsource,whichissupportedbyotherdevices,whiletheapproachtakenbyTeslaor

MercedesBenzisbasedonthefusionofcamerasandshortrangeautomotiveradars.Theselectionof

theoptimalsensorplatformisaquestionofperformance,reliabilityandcost.ThedatafromtheSA-

sensorsisusedtocreatealocalmapofthesurroundingsofthecar,whichiscomparedtoverydetailed

mapsoreven3Dmodelsoftheareawherethevehicleismoving.Thisenablesextractingtheposition

andposeofthevehiclewithmoreaccuracythanispossiblewithjustGPS-basedlocalisation.Thelocal

mapofobstaclessurroundingthecarisalsousedforreactivecollisionavoidance.[Franke,2013],

[Guizzo,2011]

ThebottompartofFigure1illustratestheshipautonomyapproachcurrentlybeingdevelopedin

AAWA.Manyexistingtechnologicalsolutionsfromautomotivedevelopmentcanbedirectly,or

throughsomeadjustment,appliedalsotoautonomousmarinenavigation.Themainquestionis

thereforenotwhethertheimplementationofautonomousshipnavigationistechnicallypossible,but

whatisthecombinationoftechnologiesandmethodsthatprovidesthelevelofperformanceand

reliabilitythatisrequiredforpracticaloperationoflargevessels,andatareasonablecost.

Thekeyaspecttosuccessfulvehicularautonomyisreliabilityandsafety.Despitealloftherecent

technologicaladvances,conclusivedemonstrationsofsufficientlyreliableautonomouscarnavigation

invaryingreal-worldconditionshavenotbeenpresented.Eventhemostadvancedandwidelytested

automotivesolutionssuchasGoogle’sautonomouscarsstillstruggletocopewithunknown

environmentsandunexpectedevents,thusrequiringhumaninterventionfromtimetotime.Even

moreimportantly,tolerancetoextremeweatherconditionsisasignificantchallenge,whichsofarhas

notbeenfullyresolved.

Publishedmarinesolutionshavesofarbeendemonstratedonsmallboatsandwithonlye.g.alimited

useofsensoryfusionandautonomyandtypicallyunderfairlyeasyweatherconditions.IntheAAWA

project,thefocusisfromthestartonharsh,butstillrealistic,conditionsandontheparticular

challengesofautonomyandremotecontrolimplementationforevenlargeocean-goingships.Thisis

madepossiblebycloseco-operationbetweenresearchersandindustrialpartners.

Fortheimplementationofautonomousnavigationandreactivecollisionavoidance,themarine

applicationpresentsbothadvantagesandchallengescomparedtootherautonomousvehicles.

Becausethespeedofashipisfairlyslow,theinterpretationofSA-sensordataandnavigation

manoeuvresdonothavetobeasfastasine.g.automotiveapplications.Theshipisalsonotconfinedto



18

e.g.anarrowroad,whichmakesavoidingothervesselseasier.Ontheotherhand,theinertiaofaship

islargeanditisnotpossibletoe.g.makeasharpturnortostopquickly.Animportantaspecttoalso

takeintoaccountisthatthenumberofautonomousshipswill,inanyfuturescenario,beordersof

magnitudelowerthanwhatisenvisionedforautonomouscars.Whileitisnotfeasibletoapplyremote

humanmonitoring(controlcentre)forbillionsofautonomouscars,shorecontrolcentresdedicatedto

autonomousshipsarefeasible.Suchcentrescanoverseetheperformanceofmultipleships,andapply

remotecontrolsifnecessary.

Thisreportwilltakeacloserlookattheavailabletechnologiesthatcanbeappliedforshipautonomy

andtheremainingchallengesaheadtoreachrequiredtechnologicalreadinessforaproof-of-concept

demonstratorbytheyear2017.

1. Autonomousnavigationofthevessel1.1. Reactivecontrolandpathplanningforcollisionfreenavigation

CollisionavoidanceforshipshasseengreatinterestafterWorldWarII,duetothedevelopmentof

radarandtherapidriseofthetrafficintheseas.Collisionavoidanceplaysamajorroleinthe

mariner’sdailyworkandbecausecriticaldecisionsofhumansarehighlysubjective,international

rulesformaritimecollisionavoidance(COLREGs)aredevelopedbyInternationalMarineOrganization

(IMO)tohelpnavigation.

Collision-freemotiontechniquescanbedividedintoeitherglobalmethods,basedonpathplanning

usingaprioriinformation,orlocalmethodswhicharebasedonreactivenavigationusingsensory

information.Inmotionplanningthepathissolvedbycomputingageometricaltrajectoryavoiding

knownobstacles,which,inreal-worlduncertainenvironments,willeasilyleadtocollision.Inreactive

navigationtherealityoftheenvironmentduringmotionistakenintoaccountusingarapidlyrepeated

perception-actionprocess.[Statheros,2008],[Pietrzykowski,2009],[Tam,2009],[Campbell,2012].



19

Planningacollisionfreepathforanautonomousmachinethroughanenvironmentcontainingstaticor

movingobstacles,inthiscaseavesselmovinginbothharbourareaandopensea,isaproblemthathas

beenextensivelystudiedduringthepastdecades.Differentsystemsrequiredifferentplanning

strategies.Also,thekinematicanddynamicconstraintsofthevesselhavetobetakeninto

considerationwhenplanningthepath,sothattheplannedmanoeuvrescanbeexecuted.Forexample,

theturningradiusofthevessellimitstheminimumturningangleallowedforthepath.Also,the

dynamicsofthevesselneedtobetakenintoaccount,i.e.thevesselturningradiusalsodependson

speedofthevehicle.Forautonomousships,alsotheenvironmentalelementsneedtobetakeninto

accountwhenplanningapath.Weatherconditionshavealsoalargeeffectontheselectionofthebest

path.Thechallengesrelatedtoreactivenavigationaremainlyduetoinstabilityoftheclosedloop

controlduetothedynamicpropertiesoftheshipandsurroundingenvironment(waves,wind,sea

currents)andingettingtheproperinformationfromtheship’ssituationalawarenesssensors.

[Statheros,2008],[Pietrzykowski,2009],[Tam,2009],[Campbell,2012],[Elkins2010].

Autonomouscarexample

ProposedmarinepipelineinAAWA

Figure1.Comparisonbetweenautomotiveandmarinenavigationpipelines.



20

Twoofthemostcommonpathplanningapproachesaregraph-basedandsampling-basedapproaches.

GraphbasedapproachessuchasA*andD*andtheirnumerousvariantshavebeenthemoststudied

algorithmsforoptimalpathplanningproblems.Themainadvantageofsampling-basedapproaches,

suchasprobabilisticroadmap(PRM)andrapidlyexploringrandomtree(RRT)andtheirvariants,is

theabilitytoeasilyincludedynamicandkinematicconstraintsofthevehicle.Forreactiveobstacle

avoidance,theseoptimalpathplanningapproachesmaynotbeefficientenough.Therefore,algorithms

suchasvelocityobstaclesarecommonlyused.[Campbell,2012],[Casalino,2009],[Lalish,2012],

[Evans,2008],[Sharma,2012],[Statheros,2008]and[Tam,2009]

1.2. AutonomousNavigationSystem(ANS)ofAAWA

InAAWA,asolutionfortheintegrationofacompleteautonomousshipnavigationarchitectureis

beingdeveloped,whichtakesadvantageofaRolls-RoyceDynamicPositioning(DP)systemdeveloped

forfutureautonomousshipsandlinksitwithanAutomaticNavigationSystem(ANS),including

SituationalAwareness(SA),CollisionAvoidance(CA),RoutePlanning(RP),andShipStateDefinition

(SSD)modulesdevelopedintheAAWAproject.Figure2showsaschematicoftheANSarchitecture.

Figure2.AutonomousNavigationSystem(ANS)architecture.

ThehighestlevelintheANSsystemistheShipStateDefinition(SSD)moduleor“VirtualCaptain”(VC),

whichcombinesinformationfromdifferentANSsub-systems(SA,DP,RPandCA),aswellasfrom

othershipautomationsystemsandtheoperatortodeterminethecurrentstateoftheship’ssystems.

Thestateoftheshipdeterminestheallowedshipoperationmode,suchasautonomous,remote-

controlorfailsafe.ThestateinformationfromtheVCisalsousedtocontinuouslyinformtheoperator

aboutthestageoftheship.



21

Dynamicpositioningsystemsallowtheshiptoautomaticallymaintainitspositionorheadingbyusing

itspropellers,ruddersandthrusters.Whencombinedwithaglobalorlocalpositioningreferencesuch

asGlobalNavigationSatelliteSystem(GNSS),andwithwindsensorsandInertialMeasurementUnits

(IMUs),theshipisabletokeepitspositioneveninroughweatherconditions.ModernDPsystems,

suchasRollsRoyceIconDP,areabletoalsomanoeuvretheshipatslowspeed.Thisallowsthe

integrationofautonomousbehaviourinshipcontrol.AstheDPsystemalreadyhasinformationofthe

ship’smanoeuvringcapabilities,itisabletocalculatewheretheshipiscantomoveinthefuture.

Thesedynamicconstraintsontheship’smovementaretransmittedtotheCAmoduletoenablemore

efficientlocalpathplanning.

RoutePlanning(RP)moduleisasoftwaremodulethatisresponsibleforplanningthepathfromstart

tofinish,viapredefinedwaypoints,whileavoidingstaticobstaclesdefinedinelectronicnavigational

chartsandfollowingshippinglaneswhenadvisable.Thismoduleiscloselyrelatedtovoyageplanning

thatisnowadaysdonebytheshipcrew.However,theRPmoduleusestheplannedvoyageas

informationwhenplanningtheactualroutefortheship.Routeconsistsofwaypoints,headingsand

speedfortheship.TheRPmoduledoesnotplanroutesinrealtimeastheCAmoduleisresponsiblefor

manoeuvresdonetoavoidobstacles.

TheCollisionAvoidance(CA)moduleisresponsibleforsafeandcollisionfreenavigation.Ituses

informationfromtheRoutePlanningmoduletofollowapaththatleadstothedestinationbutcan

deviatefromthecourseifariskofcollisiondetected.TheSAmodulesuppliesthelocalmapand

obstacleinformationthatshowsthecurrentobstaclesneartheship.TheDPmodulesuppliestheCA

modulewithanareawheretheshipisabletomanoeuvreandthuscreatesboundariesfornew

waypointsthatcanberealisticallyassigned.TheCAmodulehastwomainfunctionalities,thefirstisan

assessmentofthecollisionriskandthesecondistonavigatetheshipsafelybothintheharbourandin

theopensea.Whenacollisionriskisdetected,asuitablestateisrequestedfromtheSSDmodule,in

whichafinaldefinitionoftheshipstateismadebasedonallgivendatafromdifferentsub-systems.

Thesituationalawareness(SA)moduleoftheANSisconnectedtomultiplesensordevicesofdifferent

types.TheSAmodulefusesthesensordataandextractsrelevantinformationontheship’s

surroundingstobeusedbytheCAsystem.TheSAmodulecanalsoperformreductionofsensordata

formoreefficientoff-shipdatacommunication.TechnologydevelopmentissuesrelatedtotheSA

systemandtheshipsensorsarediscussedinSections2and3.



22

1.3. Environmentalmappingandobstacledetectionforautonomousshipnavigation

Mappingmeansthecreationofarepresentationoftheworld.Therearemultiplewaysthemapping

processcanbeperformedandwhatkindofapresentationoftheworldiscreated.Theseare

dependentontheapplication,wherethemapsareneededandwhatsensorsareusedforperceiving

theenvironment.Mapinformationisusedinforpathplanning,obstacleavoidance,andlocalisationof

theautonomousship.

Onseaandharbourarea,itispossibletousenauticalandterrainchartstoobtaininformationabout

shippinglanes,shoalsandcoastalterrain.Dynamicobstacles,suchasothervessels,aremappedby

usingtheship’ssituationalawarenesssystem,combinedwithe.g.AISdata.Manymethodshavebeen

developedforprocessingperceptiondataformodellingandrepresentinga2Dor3Dworld,to

mentionforexampleoccupancygridmaps,heightgridandQuadtreetypeofmaps.[Mooney]

Twoofthemostcommonapproachesforpresentingtheworldaretopologicalandmetricmaps.

Topologicalapproachesdescribetheconnectivityofspatiallocationsintheenvironment,whereas

metricmapsdescribetheworldthroughageometricpresentation.Topologicalmapsarebestsuited

forhigh-levelpathandmissionplanning.Metricmapscontaingeometricinformationthatisnecessary

toplanandexecutetrajectoriessafelywhileavoidingcollisions.Themappingprocesscreatesa

representationofthesurroundingworld.[Elfes,1987],[Broten2012].

Obstaclescanbepresentedaspartsofthemap,butitcanalsobebeneficialtopresentdynamic

obstaclesseparately.Objectdetectionandtrackingiscloselyrelatedtoobstacleavoidanceprocedures

andtogethertheyensurecollisionfreenavigationofthevessel.Thereareseveralmethodsdeveloped

forobstacletracking,commonlyusedareparticleandextendedKalmanfilters.Whenaseparate

presentationisusedfordynamicobstacles,usingnovelsensorfusiontechniquesandcommercialship

objecttrackingfunctionalities(ARPA),theirmovementsandactionscanbeeasiertopredict.For

example,obstaclescanhavespeedorapredefinedpath,aswellaskinematicpropertiesthatcanbe

usedtopredicttheirpositionsinthefuture.[Sinisterra,2014]



23

2. SituationalAwareness(SA)forautonomousships

2.1.Sensoringtheshipenvironment

Methodsforthefusionofmultiplesensortypes,suchasLIDARs,camerasandradarshavebeen

activelystudiedforautomotiveapplications[Herpel,2008],[Mukhtar,2015].Forexample,ashort-

rangeradarorLIDARcanprovideaccuraterange,velocityandangularmeasurementofobjects,while

cheaperandsmallercamerascanprovidebetterspatialresolutionforobjectclassification.Near-IR

(NIR)cameras,withactiveillumination,orthermalLWIRcamerascanbeusedalsofornight-time

imaging.Ontheotherhand,theuseofaradarallowsoperationalsounderdifficultweatherconditions

(e.g.heavyrainorsnow)wherethecameras(includingIR)mayfail.Thesameissuesapplyalsoto

marineSAsensors.

Themaintaskofsensorfusionistocombinethedatafromdifferentsensorsourceinsuchawaythat

optimalSAperceptionisguaranteedunderallconditionsandinallsituations.SAdataisthenusedto

maplocalobstaclestoenablereactivecollisionavoidance.

2.1.1SensortechnologiesforSituationalAwareness

Cameras

CamerasareanaturalchoiceforSA.Theyarecheap(withsomeexceptions),smallinsizeanddurable,

andcanprovideveryhighspatialresolutionwithcolourinformationforobjectidentification.True

night-visionispossiblewiththermalIRimagersandapairofcamerascanbeusedinastereoscopic

configurationfor(limited)3Dsensing.Duetothehugerangeofbothcommercialandniche

applications,cameratechnologyisstillconstantlyimproving.Thelargeexistingknowledge-baseon

visualanalysisalgorithmsprovidesmanypotentialsolutionsalsoformarineSituationalAwareness.

NormalvisualspectrumHDcamerasareseenasanimportanttechnologytobefusedwithother

sensorydata.Highspatialresolutionallowsforrecognitionofobjectsandobstacles,eitherbyahuman

remoteoperatororthroughautomatedanalysisalgorithms,andcolourinformationcanbeusedto

helptheseparation(segmentation)ofrelevantobjectsfromthebackground(seasurface).

Adisadvantageofcamerasisthemassiveamountsofdatageneratedbyhigh-resolutionsensors,

whichrequiresextensiveprocessingperformanceandhigh-bandwidthdatalinksforanalysisand

transmission.However,whenconsideringamarineSAimplementationonalargeship,the



24

requirementse.g.intermsofthesmallsizeorlowpowerconsumptionofprocessinghardware,are

muchlessstrictthanformanyotherautonomousplatforms,suchascarsoraircraft.

Visualspectrumcamerashavesomeseverelimitations:theycannotbeusedinthedark(apartfrom

detectinglights)andtheirseeingdistancedropsveryquicklyinbadweather,suchasfogorheavyrain.

BetterperformancecanbeobtainedwithcamerasoperatingintheInfrared(IR)range.Near-IR(NIR)

sensingiscommonlyusedfornight-visioninsecuritycameras,becauseNIRsignalscanbecaptured

withinexpensiveCMOS/CCDcamerasensors.This,however,requiresactiveIRilluminationofthe

scene,whichisnotpracticalforshipSA.Truepassivenight-visioncanberealizedwithLong-WaveIR

(LWIR)cameras,whicharesensitivetoIRradiationinthe8-14µmwavelengthrange.Because

thermalLWIRradiationispassivelyemittedbyallobjects,LWIRsensorscanbeusedforimagingin

totaldarkness.Duetothevaryingthermalemittanceproperties,dependingone.g.surfacematerials

andsurfacegeometry,avisuallymeaningfulimagecanbecreatedevenfromobjectsandscenes,where

theaveragetemperatureiseffectivelyuniform.AscanbeseenfromFigure3,thermalimagingcanbe

beneficialevenindaylightconditions,e.g.indifficultilluminationconditions.

Microbolometer-basedLWIRcamerasarethemostaffordablethermalimagingtechnology.

Furthermore,unlikesomeotherIRtechnologies,bolometersensorsdonotrequirecryogeniccooling,

leadingtosimpler(morerobust)camerahardware.Thedisadvantageofbolometer-basedLWIR

sensorsistheirlowresolution(typically640x480pixels,megapixelsensorsareavailablebutvery

expensive),i.e.forthesamespatialaccuracy,thefield-of-view(FOV)isnarrowerthaninanormalHD

camera,ascanbeseenfromFigure4.

Figure3.DaylightScenecapturedwithanormalcameraandathermalcamera.



25

Morerecently,Short-WaveIR(SWIR)cameratechnologyhasbecomeavailablealsofornon-military

andorscientificapplications[Stark,2015].SWIRsensorsoperateinthe1-3µmwavelengthregion,

wherethedetectedsignalisnotpassivelyemitted(thermal),butreflectedradiation.SWIRsensors

providebettervisibilitythroughhazeorfogthanvisualspectrumcamerasandtheyalsoworkwellin

verylowlightconditions,butnotintotaldarkness.IthasbeenstatedthatSWIRenablesbetter

detectionrangeunderhumidandfoggyconditionsthanLWIR[Wallace,2013].However,SWIR

technologyiscurrentlymoreexpensivethane.g.LWIR,anddoesnotimproveonthespatialresolution.

WhileIRsensorsofferbettervisibilitythanvisualrangecameras,theirperformanceisalsodegraded

inbadweather.Forexample,differentIR-bandsareattenuateddifferentlydependingonthelevelof

humidityintheatmosphere,whichcanleadtogreatlyvaryingseeingrangesdependingonweather

conditions[Beier,2004].Thisiswhyasensorsourcewhichisrobustagainstweathereffects,suchas

radar,hastobefusedwiththelessreliablecameradata.

RadarandLIDAR

Camera-basedsensing(fusionofvisualandthermalimaging)hastwosignificantdisadvantages

regardingtoSAextractioninautonomousvehicles1)insufficientweathertoleranceand2)lackofan

easywaytoextractobjectdistanceinformation.

Acombinationoftwomonocularcamerascanbeusedtoimplementstereoimaging,i.e.createa3D

mapofthevisualscenethroughdisparitymappingbetweentwoimages.Thedrawbackofstereo

imagingisthecomputationalcomplexityrelatedtolargeamountsofimagedataappliedtostereo

matchingalgorithm.Also,thechoiceofcamerabaseline,i.e.thephysicalseparationofthetwosensors,

effectivelysetsconstraintsonthedistanceresolvingcapabilityofthesystem.Muchbetter

performancecanbeobtainedbyusingactivesensortechnologies,suchasradarorLIDAR.

Figure4.Fusionofvisuallightandthermalimagingatnight.



26

Inmaritimeapplicationstheuseofradarhasalonghistory.Therefore,severalradarsystemsuppliers

canbefoundinthemarketforobstacledetectionandmapping.Radarcapabilityisinfluencedbythe

operatingfrequencybandoftheradar,sothattypicallyhigherfrequenciesofferbetterangleandrange

resolution.Thereisawidevarietyofradarsinthemarket,intendedfordifferentpurposes,having

specificcarrierfrequencies,bandwidths,transmitdurations,waveforms,antennasetc.Typically,

marineradarsaremicrowaveradarsusingS-orX-bands,whicharerobustindifferentweather

conditions.[Heuel,2013]

However,theresolutionoftraditionalmarineradarmaynotbesufficientforreactivecollision

avoidance.Forexample,consideringanautonomousshipinaharbourareaorapproachingthedock,

theresolutionoftheradarintheverynearfield,i.e.somehundredsofmeters,needstobegood

enoughtobeabletodetect,andmaybealsotrack,evensmallstationaryandmovingobjects.NewKa

andW–bandradars,originallydevelopedforautomotiveapplications,couldbebeneficialin

autonomousshipapplications,especiallyforverycloserangeobstacledetection.Theyoffermuch

betterangularanddistanceresolutionthantraditionalshipradars,atthecostofreducedrange.These

newtypeofradarstogetherwithmodernS-andX-bandradarsandseveraldifferenttypeofcameras

areexploitedinthedevelopmenttoenablenear-fieldreactivecollisionavoidance,aswellas

autonomousnavigationine.g.harbourareas.[Skolnik,2008],[Seliga,2010]

LIghtDetectionAndRanging,LIDAR(orLAserDetectionAndRanging,LADAR)isascanninglaser

sensortechnology,whichcanprovideveryaccuratedistancemeasurements.Multichanneldevices

(e.g.witha64laserarray),suchasthoseusedinGoogle’sautonomoustestcars,cancreateavery

detailed3Dmapofthesurroundingsofthevehicle.LIDAR-basedmarinenavigationhasbeen

proposedanddemonstratede.g.in[Jimenez,2009],[Pastore,2010]and[Halterman,2010].

Figure5.Left:radarviewofobject.Right:3DLIDARscanningdata



27

OnepossibledisadvantageofLIDARisthatitusesrapidlymovingmechanicalcomponentsforthe

scanningoperation,whichcouldbepronetomalfunctions,especiallyoverlongerperiodsoftimeina

harshmarineenvironment.BecauseLIDARemploysalaserbeam(typicallyapulsedIRlaser),its

rangeandaccuracyisalsoaffectedbyadverseweather,suchasheavyfog,rainandsnow,similarlyto

IRcameras.

2.2.Sensordatafusionandprocessing

Inpracticallyallfieldsofvehicleautonomy,utilisationsensorfusionhasbeenseenasthekeyfor

achievingsufficientsituationalawarenessreliability.Eachseparatesensortypeexhibitsparticular

weaknessesandlimitationsundersomeconditions(weather)ordetectionsetups(range,fieldofview,

identification).Also,bothfalsepositiveandfalsenegativedetectionscanneverbecompletely

preventedforasinglesensor;optimisingoneleadstoatradeoffagainsttheother.Bycombiningthe

capabilitiesofmultiplesensormodalities,individualerrorsandweaknessescanbeaveragedoutand

betteroverallperformancecanbereached.Table1roughlycomparesdifferentpotentialsensortypes

intermsofperformanceaspectsrelevanttomarineSA.

Basedonthereviewofexistingsolutionsreportedinvariousfieldsofautonomyandthetestingof

sensortechnologiesalreadydoneinAAWA,thefusionofdifferenttypesofradarsandvisualsensors,

includingthermalIRcamerasisseenasafeasiblesolutionformarinesituationalawareness.Radars

enableeasytrackingoftargetdistanceandcanprovidetherequiredtolerancetobadweather.The

applicationofnewhigh-GHzradarsdevelopedforautomotivecollisionavoidancecanalsoprovide

sufficientobjectdetectionaccuracyforrangeswhicharetooshortandtooinaccurateforconventional

Table1.ComparisonofdifferentmarineSAsensors.

VisualHDcameras

IRcameras ShipradarShort-rangeradar

LIDAR Sound

SpatialAccuracy ++ + -- - ++ --

Fieldofview + - ++ - + ++Distancemeasurement - - ++ ++ ++ --Objectidentification ++ + -- -- + +24H,allweatheroperation -- + ++ ++ +(?) -(?)Computationalloadofanalysis -- - ++ ++ -- +Marinerobustness ++ ++ ++ +(?) (?) (?)

Price ++ - +- ++ -- +



28

shipradar.Camerascanaugmentradardatabyprovidingmoredetailedinformationondetected

objects.Furthermore,thefusionofcamerasandradarcanalsoincreasedetectionrobustness.Thermal

IRcamerascanseeobjectsalsointotaldarkness,whilecolourinformationfromnormalHDcameras

canbeusedforsegmentingobjectsinthewater.Asignificantconsiderationisalsotheaffordabilityof

thesensorysystem.Whilemanyhigh-endsensordevicescouldprovideavaluableadditiontotheSA-

system,asolutionwhichtakesadvantageofaffordabletechnologyshouldbepreferred.Inthissense,

thecombinationofradarandimaginghasanadvantageovere.g.LIDARsensors.

Soundsignalling(e.g.byhorns)isalsoanintegralpartofthecurrentmaritimenavigationprocess.

Therefore,inordertorealiseaSAsystemwhichisatleastascapableasahumancrew,soundcapture

anddataanalysisshouldalsobeincluded.Whileloudandclearsoundssuchase.g.hornsandwhistles

couldbefairlyeasilydetected,theirsourcesshouldalsobeaccuratelylocalisedrelativetotheship,to

helpreactivecollisionavoidance.Thisrequiresmorethanjustcaptureanddetectionofsound,the

soundsourcealsohastobespatiallylocalised,e.g.viaanarrayofmicrophones,andthesounddata

fusedwithothersensormodalities.Soundsensorscouldalsobeappliedinamoregeneralmannerin

theSAsystem,fordetectingandidentifyingothervesselsbythesoundsthattheynormallyemit.

Sensordataprocessing

Themostcomputationallyintensivepartofasensorfusionpipelineistheanalysisofdataprovidedby

cameras.Theoutputcreatedbyradarsisverysparse(objectswithsomenoise)andthereforemuch

easiertoprocess.Animportantpartofimagedataprocessing,isthesegmentationoftheinputdata.

Highresolutionvideocamerasprovidemassiveamountsofdata,mostofwhichisirrelevantforthe

processofobjectdetectionandimagecontentunderstanding.Thefirststepinanimageanalysis

processisthereforetosegmenttherawinputdata,i.e.toremoveallinformationwhichisnotrelevant

totheparticulartask(background)fromthosefeaturesandobjectswhichshouldbedetected

(foreground).Onthereducedamountofimagedata,morecomplexanalysisalgorithmscanthenbe

appliedforspatialandtemporalobjecttrackingandobjectclassification.Forexample,amarinescene

canbeassumedtoalwaysconsistofthreedifferentcoarseregions:waterinthebottompartofthe

image,skyinthetoppartandahorizonareainthemiddle.Byfindingthehorizonline,alargepartof

theimagedatacanbediscardedfromfurtherprocessing.Sensorfusioncanbeusedtomakethe

processeasierbyusingcluesfromothersensormodalitiestohelptheimageprocessingpipeline.For

example,thedetection(orlackthereof)ofobjectsintheviewofaradarcanbeusedtoguidetheimage

segmentationalgorithmtofocusmoreonpotentialobjectareasandfalsedetectionsfromimagedata

canbediscardedifsuitableconfidencebasedonradardataisavailable.



29

Datafrommultiplesensorscanbefusedindifferentways,asillustratedinFigure7.Low-levelfusionis

performedontherawornearlyunprocesseddatafromdifferentsensors,whileinhigh-levelfusion,

theseparatedatastreamsareprocessedindividuallyandthedetectionsfromdifferentsensorsare

combinedonobjectlevel.Theuseoflow-levelfusionismorenaturalbetweentwodifferentcamera

types,suchasvisualandthermalsensors,whilethefusionbetweencamerasandradarcanbemore

naturallyimplementedonahigherobjectlevel.Inpractice,themostefficientwaytoimplementsensor

fusionbetweenmultiple(>2)differentsensormodalitiesisprobablyacombinationofbothlow-level

andhigh-levelfusionapproaches.

Inamarinesensorfusionprocess,radarcanbeusedtoprovidebearinganglesanddistancesfor

variousobjectsinthescene.Thisinformationcanthenbemappedtocorrespondingobjects

segmentedfrommultiplecameradata,toextractmoredetails.Thepresenceofthesameobjectin

multiplesensordataprovidesamorerobustdetectionthanasinglesensorsource,whichcanalways

providenoisyorincompletedata.Frame-to-frameanalysisresultsoftencontaintemporalnoise,with

objectsbeingsometimeslostduetoanalysisuncertainty.Spatialandtemporalobjecttrackingcanbe

appliedtoprovideacontinuoussituationalawarenessforreactivecollisionavoidance.

Inordertoreachbestpossibleautonomousnavigationreliability,allotheravailabledatasources

whichcanhelptheshipnavigationandcollisionavoidanceprocessshouldalsobefusedwithonboard

sensordata.TheseincludealreadycommonlyusedtechnologiessuchasGPS,AIS,ARPAandECDIS,the

outputsofwhichcanbefusedwiththeextractedsensordataviahigh-levellocalandglobalmap

representations.

Figure6.Horizondetectionandobjectsegmentation.



30

3. Off-shipcommunication

Thecapabilityforremotehumaninteractionandcontrolhastobeenabledforsituations,whichthe

shipautonomycannotresolveorisnotallowedtohandlebyitself.RelayingtheSAinformation

gatheredbytheship’ssensorstoaremoteoperatormayrequirethetransferofsignificantamountsof

data.Duetopracticallimitationsone.g.satellitecommunicationsatopensea,thesameamountof

bandwidthmaynotbeavailableatalltimes.Methodsforreducingtheamountofsensordataonlyto

whatisabsolutelyneededforthehumanoperatortoperceivetheenvironmentoftheshipneedstobe

considered.Alsoissuessuchasdatasecurity(intentionaltampering)andlinkreliabilityshouldbe

addressedandthepossibilitiesofusingmultiplealternativecommunicationnetworks(satellite,VHF,

4G)dependingonavailabilityandperformanceneedsshouldbeexamined.

TransmissionofHDvideofromtheshiptotheshorecontrolcentreisnotrequiredallthetime.Itmay

berequiredonlywhensomethingunexpectedthatrequirestheattentionoftheshorecontrolcentre

happens.Suchasituationcouldbeforexampledetectinganobstaclewhichrequireshuman

identification,orasituationinwhichtheshipisunabletocalculateareliableavoidancemanoeuvre.It

isassumedthatformostofthetimeintheopenseas,theautonomouscontrolsystemisabletohandle

thesituationwiththehelpofthesensorysystemsonboard(collisionavoidance,objectdetectionetc.).

Thus,mostofthetime,veryminimalamountofoutbounddata,suchasshipstateinformationand

reducedsensordataisrequired.Ontheotherhand,sufficientlyhightransmissioncapacityshouldbe

availablewhenneededonshortnotice.

Theamountofdatatobetransmittedgrowsquicklyasmoresensorsareaddedtothesystem,

especiallywithhighresolutionvideo.Reductionoftheframe-rate,lowerimageresolutionandefficient

videocompressionhavetraditionallybeenappliedforremotemonitoringoverlow-bandwidth

datalinks.However,toevenfurtherreducedatatransferrequirements,thesensorinputscanbe

segmentedwiththeonboardSAprocessingsystemtoextractonlytheminimalamountofdata,which

canstillbesufficientforhumanunderstandingofthescene.Foreground/backgroundsegmentation

performedbytheship’sSA-system,enablestransmissionofonlycertainrelevantfeatures,objectsor

regionsofinterest(ROI),asillustratedinFigure8.Ahumanoperatorcould,atleastinnon-critical

conditions,extractsufficientsituationalawarenessfromverysparsesegmentedimagefeatures,which

canrequirethetransferofverylittledata.



31

Ontheopensea,themainmeansofcommunicationisviasatellite,however,satellitecommunication

canbedisturbedbyweatherconditions.Theamountofattenuationcausedbye.g.heavyrainis

dependentonthefrequencybandemployedbythesatellitenetwork.Forexample,fadingismuch

moresevereatKa-bands(above20GHz)thanattheL-band(1to2GHz)[Qingling,2006].Thismeans

thatsevereweathermaydegradetheperformanceoflinksoperatingatKa-bands.However,

combiningaKa-bandsystemwithe.g.alessweathersensitiveL-bandnetwork,ashasbeendonein

theInmarsatGlobalXpresssystem,reducestheriskoflosingallcommunicationseveniftheKa-band

systemwouldbenon-operational.TheInmarsatsystemallowsdynamicswitchingbetweenthetwo

satellitetypeswithoutusereffort.However,thelowercapacityofferedbytheL-bandsatelliteshasto

betakenintoaccountwhenallocatingbandwidthtooff-shipcommunication.

Inthefuture,theremaybealargenumberofautonomousvesselsinthesamesatellitebeamor

cellularnetworkcellarea.Asthetotalbandwidthwithinacertainbeamorcellareaissharedbetween

allusers,ashortageofbandwidthmaybecreatedifmanyvesselssimultaneouslyrequirehigh

bandwidth,forexampleforHDvideotransmission.Thisproblemcouldbeleveragedbyforming

swarmsorfleetsofvesselswhereoneshipwouldbetheleader.Inthisway,communicationtoashore

controlcentrecouldbecoordinatedviatheleadshipwithlineofsightship-to-shipcommunication.In

thiswayitcouldbepossibletooptimisetheuseofsatellitebandwidthinacertainareabyreducing

theneedforalltheshipsintheswarmtocommunicatewiththeshorecontrolcentresimultaneously.

Figure7:Imagesegmentationforreducingdatarates.Top-left:Original(1900x1080px)withsegmentedobjects.Top-right:1-bitedgedata.Bottom-left:1-bitobjectdata.Bottom-right:Regionofinterestcaptureofobjectarea(241x145px).



32

Possibleeffectsofweatherormulti-usercongestiononcommunicationperformanceshouldbe

consideredcarefullywhenimplementingthecontroland“intelligence”ofthewholeautonomysystem

throughthe“VirtualCaptain”.Difficultsituationsmayariseifpoorweathersimultaneouslycauses

reductionofSA-systemcapability,requiringmoreshorecontrolinterventionordecisionmaking,and

areductionindatalinkcapabilityrequiredtotransfersensordatafromtheship.Correctbehaviours

andprecautionsforsuchsituationsshouldbedefined.TheseissuesareaddressedintheANS

architecturedevelopmentinAAWA,throughtheVirtualCaptainandtheshipstatedefinitiondiscussed

earlier.

Referencesintext

1. [Beier,2004]K.Beier,H.Gemperlein:“Simulationofinfrareddetectionrangeatfogconditions

forEnhancedVisionSystemsincivilaviation”,AerospaceScienceandTechnology8,pp.63–71,

Elsevier,2004.

2. [Broten,2012]G.Broten,D.MackayandJ.Collier,"Probabilisticobstacledetectionusing21/2

Dterrainmaps",inComputerandRobotVision(CRV),NinthConferenceOn,pp.17-23,2012.

3. [Cambell,2012]S.Cambell,W.Naeem,G.W.Irwin.“Areviewonimprovingtheautonomyof

unmannedsurfacevehiclesthroughintelligentcessionavoidancemanoeuvres”,Annual

reviewsinControl,2012.

4. [Casalino,2009]Casalino,G.;Turetta,A.;Simetti,E.,"Athree-layeredarchitectureforrealtime

pathplanningandobstacleavoidanceforsurveillanceUSVsoperatinginharbourfields,"in

OCEANS2009-EUROPE,vol.,no.,pp.1-8,11-14May2009

5. [Elfes,1987]A.Elfes,"Sonar-basedreal-worldmappingandnavigation,"IEEEJournalOf

RoboticsandAutomation,vol.3,pp.249-265,1987.

6. [Elkins,2010]L.Elkins,D.Sellers:“TheAutonomousMaritimeNavigation(AMN)Project:Field

Tests,AutonomousandCooperativeBehaviours,DataFusion,Sensors,andVehicles”,Journal

ofFieldRobotics27(6),pp.790-818,2010.DOI:10.1002/rob.20367

7. [Evans,2008]J.Evans,P.Patrón,B.Smith,D.M.Lane,(2008),“Designandevaluationofa

reactiveanddeliberativecollisionavoidanceandescapearchitectureforautonomousrobots.“

AutonomousRobots,24(3),247-266.2008.

8. [Franke,2013]U.Franke,D.Pfeiffer,C.Rabe,C.Knoeppel,M.Enzweiler,F.Stein,R.Herrtwich:

“MakingBerthaSee”,Proc.ofthe2013IEEEInternationalConferenceonComputerVision

Workshops,pp.214-221,2013.

9. [Guizzo2011]E.Guizzo:“HowGoogle’sSelf-DrivingCarWorks”,IEEESpectrum

(http://spectrum.ieee.org/),18Oct2011.



33

10. [Halterman,2010]R.Halterman,M.Bruch:“VelodyneHDL64ELIDARforUnmannedSurface

VehicleObstacleDetection”,SPIEProc.7692:UnmannedSystemsTechnologyXII,Orlando,FL,

April2010.

11. [Heuel,2013]HeuelS.,RadarWaveformsforA&DandAutomotiveRadar,whitepaper,2013.

12. [Herpel,2008]T.Herpel,C.Lauer,R.German,J.Salzberger:“Trade-OffbetweenCoverageandRobustnessofAutomotiveEnvironmentSensorSystems”,Proc.ofInternationalConferenceon

IntelligentSensors,ISSNIP2008,pp.551-556,2008.

13. [Jimenez,2009]A.R.Jimenez,F.Seco:“AShort-RangeShipNavigationSystemBasedonLadarImagingandTargetTrackingforImprovedSafetyandEfficiency”,IEEETransactionson

IntelligentTransportationSystems,Vol.10,No.1,pp.186-197,2009.

14. [Lalish,2012]E.Lalish,K.A.Morgansen,K.A.(2012).“Distributedreactivecollision

avoidance“,AutonomousRobots,32(3),207-226.2012.

15. [Mooney]D.Mooney,“MetricPathPlanning”,http://www.eecs.ucf.edu/~gitars/cap6671/Papers/metric_planning.pdf

16. [Mukhtar,2015]A.Mukhtar,L.Xia,T.B.Tang:“VehicleDetectionTechniquesforCollisionAvoidanceSystems:AReview”,IEEETransactionsonIntelligentTransportationSystems,Vol.

6,Issue5,pp.2318–2338,2015.

17. [Pastore,2010]T.J.Pastore,A.N.Patrikalakis:“LaserScannersforAutonomousSurfaceVesselsinHarborProtection:AnalysisandExperimentalResults”,Proc.ofthe2010International

WatersideSecurityConference,pp.1-6,2010.

18. [Pietrzykowski,2009]Z.Pietrzykowski,Z.,&Uriasz,J.(2009).Theshipdomain-acriterionof

navigationalsafetyassessmentinanopenseaarea.JournalofNavigation,62(1),93.2009.

19. [Qingling,2006]Z.Qingling,“RainAttenuationinMillimeterWaveRanges”,7thInternationalSymposiumonAntennas,Propagation&EMTheory,ISAPE'06,pp.1-4,2006.

20. [Seliga,2010]T.A.Seliga,(2010,March),“HighResolutionW-BandRadarDetectionandCharacterisationofAircraftWakeVorticesinPrecipitation“,InWakeNet3-Europe/Greenwake

dedicatedworkshopon”Wakevortex&windmonitoringsensorsinallweatherconditions.

2010.

21. [Sharma,2012]R.Sharma,J.B.Saunders,R.W.Beard,(2012),“Reactivepathplanningfor

microairvehiclesusingbearing-onlymeasurements“,JournalofIntelligent&RoboticSystems,

65(1-4),409-416.2012.

22. [Sinisterra,2014]Sinisterra,A.J.;Dhanak,M.R.;vonEllenrieder,K.,"Stereovision-basedtarget

trackingsystemforanUSV,"inOceans-St.John's,2014,vol.,no.,pp.1-7,14-19Sept.2014

23. [Skolnik,2008]MerillSkolnik“RadarHandbook”(3rdedition),McGraw-HillEducation;3

edition(February12,2008),ISBN-13:978-0071485470.



34

24. [Stark,2015]B.Stark,M.McGee,Y.Chen:“ShortWaveInfrared(SWIR)ImagingSystemsUsing

SmallUnmannedAerialSystems(sUAS)”,Proc.ofthe2015InternationalConferenceon

UnmannedAircraftSystems(ICUAS),pp.495-501,2015.

25. [Statheros,2008]T.Statheros,G.Howells,K.M.Maier."Autonomousshipcollisionavoidance

navigationconcepts,technologiesandtechniques."Journalofnavigation61.01(2008):129-

142.

26. [Tam,2009]C.K.Tam,R.Bucknall,A.Greig,“ReviewofCollisionAvoidanceandPathPlanning

MethodsforShipsinCloseRangeEncounters“,JournalofNavigation/Volume62/Issue03/

July2009,pp455-476Copyright©TheRoyalInstituteofNavigation2009

27. [Wallace,2013]J.Wallace:“IRIMAGING:SWIRcamerasspotunwantedactivitiesthroughfogandatnight”,LaserFocusWorld(online),Vol.49,Issue10,2013

http://www.laserfocusworld.com/articles/print/volume-49/issue-10/world-news/ir-

imaging-swir-cameras-spot-unwanted-activities-through-fog-and-at-night.html



35

LegalitiesHenrikRingbom,AdjunctProfessorinMaritimeLawandtheLawoftheSea

DepartmentofLaw,ÅboAkademiUniversity

FelixCollin,ProjectResearcher,FacultyofLaw,UniversityofTurku

MikaViljanen,ActingProfessorofCivilLaw,FacultyofLaw,UniversityofTurku



36

LegalImplicationsofremoteandautonomousshipping

1. Introduction

Maritimelawisafunctionaltermusedfordescribingawholerangeoflawsandotherlegalsources

thatgovernthelegalframeworkrelatedtoshipsandtheiroperation.Itincludesavarietyofdifferent

legalsystems,rangingfrominternationallawtoregionalandnationalrulesanddowntolocalrules.It

coversissuesofpublicconcerns,suchassafety,securityandenvironmentalprotectionaswellascivil

lawmatters,suchascontractsofcarriage,liabilityandcompensationfordamage,salvageandrules

relatedtomarinerisksandinsurance,tonamebutafew.

Theprospectofunmannedshipsaddressesaveryfundamentalfeatureinshipping–theroleofthe

masterandcrewonboardaship–andwillhenceaffectamultitudeoflawsandregulationacrossthe

wholerangeofmaritimelaw.Anefforttosummarisethedifferentlevelsandtypesofrulesconcerned

ismadeinthetableintheannex.

Thefocusofthispaperisontheinternational(global)rules.Threemainkindsofsuchrulesneedtobe

distinguished.First,therearejurisdictionalrules,whichlaydownstates’rightsandobligationstotake

measureswithrespecttoships.Thesearemainlylaiddowninthe1982UNConventionontheLawof

theSea(UNCLOS),whichisdiscussedinsection2.Second,thetechnicalrulescoveringsafety,

environmentandtrainingandwatchkeepingstandardsetc.arediscussedinsection3.Theyareusually

adoptedbyspecialisedUNagencies,suchasnotablytheInternationalMaritimeOrganization(IMO).

Third,aseriesofinternationalruleshavebeenestablishedinthefieldofprivatelawtoharmonise

issuessuchasshipowners’civilliabilityforpollution,collisionsorcargo-relatedlossesandhowsuch

claimsmaybeenforced.Theserulesarenotascompleteorwidelyratifiedasthepubliclaw

conventionsdiscussedinsections2and3andmaythereforebesubjecttogreaternationalvariation.

Themainrelationshipsoftheserulestoautonomousshippingarediscussedinsection4.

2.LawoftheSea2.1General

Thelawoftheseadealswiththerightsandobligationsofstatesovertheseas.Asfarasshippingis

concerned,thekeyissuesaddressedbythisbodyoflawinclude:towhatextentcanshipsnavigatein

differentseaareas;whatobligationsdostateshaveovershipsflyingtheirflag;andwhatrightsdo

otherstateshavetointerfereinthenavigationofshipsindifferentseaareas?



37

Today’slawoftheseagoverningnavigationismorestablethaneverbeforeinhistory.The

‘ConstitutionfortheOceans’,UNCLOSenjoysawidespreadformalacceptanceworldwide(167

contractingparties)anditsprovisionsconcerningnavigationalrightsanddutiesarewidelyaccepted

asrepresentingcustomarylaw(andhenceapplytonon-partiesaswell).Theconventionlaysdownthe

rulesonestablishmentanddelimitationofmaritimezonesandincludesdetailedrulesonstates’rights

andobligations,differentlyforeachzone.

Afirst–andfundamental–questiontoberesolvediswhethershipswithoutacrewonboardare

‘ships’or‘vessels’withinthemeaningoftheconventionatall.Thetwotermsareusedinterchangeably

inUNCLOS,butneitherisdefined.Itdoes,however,followfromthenatureoftheactivitiescarriedout

bythelarge,self-propelled,cargo-carrying,commercially-operatedunmannedshipsofinteresthere

thattheyprobablywillhavetoberegardedasvessels/shipsbyvirtueoftheirsize,featuresand

functions.Existinginternationalconventionsthatdefinethetermshipdonotincludereferencesto

crewingandatnationallevel,too,thedefinitionofashipisusuallydisconnectedfromthequestionof

whetherornottheshipismanned.1Itwouldalsoseemunjustifiedthattwoships,onemannedandthe

otherunmanned,doingsimilartasksinvolvingsimilardangerswouldnotbesubjecttothesamerules

thathavebeendesignedtoaddressthosedangers.

Fromtheassumptionthatunmannedshipsare‘ships’and’vessels’withinthemeaningofUNCLOS

followsthattheyaresubjecttothesamerulesofthelawoftheseaasanyordinarilycrewedship.The

sameobligationsapplytounmannedshipsandtheirflagstateswithrespecttocompliancewith

internationalrules.Ontheotherhand,theyalsoenjoythesamepassagerightsasothershipsand

cannotberefusedaccesstootherstates’watersmerelybecausetheyarenotcrewed.

2.2FlagStateJurisdiction

Flagstatejurisdictionrepresentsthetraditionalcornerstoneoftheregulatoryauthorityoverships.

UNCLOSestablishesthatallstateshavearighttosailshipsflyingtheirflagandtofixtheconditionsfor

grantingnationalitytoships(Articles90and91(1)).However,theconventionalsoincludesanumber

ofdetaileddutiesforflagstates.

Everystatehastheobligationto“effectivelyexerciseitsjurisdictionandcontrolinadministrative,

1 Study ’Liability for Operations in Unmanned Maritime Vehicles with Differing Levels of Autonomy’, performed by the University of Southampton on behalf of the European Defence Agency’s Group on Safety and Regulation of Unmanned Maritime Systems (SARUMS), , 2016, Part 1, Chapter 3 (unpublished).



38

technicalandsocialmattersovershipsflyingitsflag”(Article94(1)),includingto“assumejurisdiction

underitsinternallawovereachshipflyingitsflaganditsmaster,officersandcrewinrespectof

administrative,technicalandsocialmattersconcerningtheship”(Article94(2)(b)).Theflagstateshall

also“takesuchmeasures…asarenecessarytoensuresafetyatseawithregard,interalia,to…the

manningofships,labourconditionsandthetrainingofcrews,takingintoaccounttheapplicable

internationalinstruments”(Article94(3)(b)),includingmeasuresnecessarytoensure“thateachship

isinthechargeofamasterandofficerswhopossessappropriatequalifications,inparticularin

seamanship,navigation,communicationsandmarineengineering,andthatthecrewisappropriatein

qualificationandnumbersforthetype,size,machineryandequipmentoftheship”(Article94(4)(b)).

Whenadoptingthesemeasureseachflagstateisrequired“toconformtogenerallyaccepted

internationalregulations,proceduresandpracticesandtotakeanystepswhichmaybenecessaryto

securetheirobservance”(Article94(5)).

UNCLOS,inotherwords,avoidstheneedtoformulatemorepreciseobligationsofflagstatesby

referringtoanabstract,andcontinuouslychanging,setofinternationalrulestobedeveloped

elsewhere.Inthiswayitavoids‘freezing’therequirementsatagivenpointintimeoratagiven

technicallevel,whilestillpreservingtheinternationalcharacteroftherulesinquestion.Themore

preciseextentofflagstates’obligationsishencelefttobedevelopedbytheIMOinparticular.

2.3PortandCoastalStateJurisdiction Whiletheflagstate’sjurisdictionappliesirrespectiveoftheship’slocation,otherstates’parallel

jurisdictionoverthesameshipdependsonthemaritimezoneconcerned.Thecoastalstate’sauthority

overaforeignshipincreaseswiththeproximityoftheshiptoitsshores.

Iftheshipisvoluntarilypresentinoneofitsportsorinternal waters,thecoastal/portstatehas

broadjurisdictionoverforeignships.Internalwatersformpartofthesovereigntyofthestate(Article

2)andintheabsenceofspecificlimitations,thejurisdictionoverforeignshipsinthisareaistherefore

complete.Moreover,shipshavenogeneralrighttoaccessforeignportsandtheportstate’swide

discretiontoplaceentryconditionsforforeignshipsiswidelyacknowledged,includinginUNCLOS

Articles25(2),211(3)and255.Inotherwords,aportstatemay(unlessithasacceptedspecific

obligationstothecontrary)refuseunmannedshipsaccesstoitsportsorinternalwaters,provided

thattherefusalcomplieswithcertainmoregeneralreasonablenesscriteriathatexistingeneral

internationallaw,suchasnon-discrimination,proportionalitybetweenthemeasureanditsobjective

andthattheprohibitiondoesnotconstituteanabuseofright(Article300).Thismayturnouttobea



39

significantlimitationofthefreeofmovementofunmannedships,butthepotentiallimitationisbyno

meansuniquetounmannedships.

Withrespecttoshipspassingthroughitsterritorial sea(whichmayextendupto12nauticalmiles

fromthecoastline/baseline),therightsofcoastalstatesaremorelimited.Underalongstanding

principleofthelawofthesea,allshipsenjoyarightof‘innocentpassage’throughotherstates’

territorialseas.Passageisdeemedtobeinnocentaslongasitisnot“prejudicialtothepeace,good

orderorsecurityofthecoastalstate”(Article19(1)).Alistofactivitiesthatmeetthosecriteriaisgiven

inArticle19(2),butasthelistfocusesonships’activities(suchasuseorthreatofforce,military

activities,fishingactivitiesorwilfulandseriouspollution)questionsrelatedtoaship’smanningwill

notassuchrenderpassagenon-innocentunderthewordingofUNCLOS.

Regardingthecoastalstate’slegislativejurisdiction,Article21(2)providesthatastatemaynotimpose

itsnationalrequirementsontheconstruction,design,equipmentormanningofforeignshipsinits

territorialsea,unlessthoserequirementsaregivingeffectto“generallyacceptedinternationalrules

andstandards”(Article21(2)).Independentlyofwhatlawsthecoastalstatehasadopted,itmaynot

“imposerequirementsonforeignshipswhichhavethepracticaleffectofdenyingorimpairingthe

rightofinnocentpassage”(Article24(1)(b).Therightofinnocentpassageextendstoshipsthatare

deemedtoposeaparticularriskforthecoastalstate,suchasnuclear-poweredshipsandships

carryingnuclearorotherinherentlydangerousornoxioussubstances(Article23).

Theareasofacoastalstate’sterritorialseawhichformspartofa‘strait usedforinternational

navigation’ aresubjecttoevenmorelimitationsforcoastalstates(andcorrespondinglystronger

passagerightsforships).Therearedifferentkindsofsuchstraits,butmanyofthemostimportant

straitsthatarecompletelycoveredbytheborderingstraits’territorialseas,suchastheStraitsof

DoverandMalacca,aresubjecttotheregimeof‘transitpassage’,whereships’rightof(continuousand

expeditious)passagearegrantedandmaynotevenbetemporarilysuspendedbytheborderingstates

(Articles37-44).Manyotherimportantstraits,includingtheDanishandtheTurkishStraits,are

governedbylong-standinginternationalconventionswhichguaranteethenavigationalrightsof

foreignships(Article35(c)).

Thejurisdictiontoprescribenationalrequirementsisobviouslyevenmorelimitedwithrespectto

shipssailingintheexclusiveeconomiczone(EEZ),whichmayextendbeyondtheterritorialsea,

uptoamaximumof200nmfromthecoastline/baseline.Inthiszonefreedomofnavigation(forall

states)applies,subjecttohavingdueregardtotheinterestofotherstates(Article58).Themostexpress

prescriptivejurisdictionofcoastalstatesoverforeignshipsintheEEZconcernslawsaimingatthe



40

protectionofthemarineenvironmentandevenhere,coastalstates’jurisdictionislimitedto

prescribingrulesthatgiveeffecttointernationalruleswhileenforcementmeasuresexclude

interferinginthepassage,saveforthemostseriouscasesofpollutionanddamage(Articled211(5)

and220).

Inseaareasthatliebeyondthejurisdictionofanycoastalstate,thehighseas,thestartingpointis

thattheflagstatealonehasjurisdictionovertheship.Anumberofexemptionstothismainruleexist,

butnoneofthemisrelevantforthequestionofnavigationalrightsofunmannedships.

2.4OtherrelevantprovisionsinUNCLOS

Apartfromthejurisdictionalprovisions,certainotherUNCLOSprovisionsmayturnouttobe

problematicforunmannedships.TheobligationsetoutinArticle94(4)(b)thateachshipneedsto

havea(properlyqualified)masterandacrewwasalreadymentionedabove.Whilethisrequirement

mayarguablybemetincaseofremotelyoperatedshipsitislessobvioushowafullyautomatedship

wouldqualify.Sinceunmannedshippingoperationswilloftenrepresentamixbetweendifferent

degreesofautomation,dependingonseaareas,trafficdensityetc.furtherclarificationsofthis

obligationmaybeneeded,atleastatthelevelofthe‘generallyacceptedinternationalrules’.

AnotherUNCLOSprovision,whichpresumesacrewonboardistheobligationofthemastertorender

assistancetopersonsindangerordistressaccordingtoArticle98(1)(asspecifiedinSOLAS

RegulationV/33).Thecommunicationpartofthedutycanpresumablybemetbyremotelyoperated

shipswithrelayedradiocommunications,butitislessclearhowphysicalassistancecanberendered

byashipwithoutacrewonboard.Thedutiesincludequalificationsbyreferenceto“insofarashecan

dosowithoutseriousdangertotheship”or“insofarassuchactioncanbereasonablyexpectedof

him”whichwillprobablyreducetheextentobligationsforunmannedships,astheavailableoptions

willbefewer.However,theabsenceofacrewdoesnotinitselfdoawaywiththedutytoprovide

assistancetotheextentnecessaryandreasonable.

3.Technicalrequirements3.1General

IMOalonehasadoptedmorethan50internationalconventionsandprotocolsaimedatharmonising

rulesforinternationalshipping.Mostoftheserulesarelaiddownintheformofobligationsimposed

onships’flagstateadministrations.Itisprimarilyfortheindividualstatespartiestotheconventions



41

ensurethateachshipflyingtheirflagisboundbyandcomplieswiththerules.Acertificatewhereby

theadministrationconfirmscomplianceisoftenrequiredandthiscertificateshallbeacceptedby

otherstatesasifitwereissuedbythemselves.Inaddition,theIMOrulesfrequentlyincludea

possibilityforportstatestoverifythatshipsthatenterintotheirportsinrealitycomplywiththe

requirementsand–ifnot–totakecorrectivemeasures,includingdetainingtheshipifnecessary.

SinceitisnotpossibletocoverallIMOconventionshere,aselectionofthemostimportant

instrumentswithimplicationsforcrewshasbeenmadehere.Theselectedconventions(SOLAS,

MARPOL,STCWandCOLREGs)areallwidelyratifiedamongtheworld’s(flag)statesandhence

applicableworldwide.TheMaritimeLabourConvention(MLC),whichwasdevelopedbythe

InternationalLabourOrganisation(ILO)in2006,hasalreadybeenratifiedbymorethan70states.

3.2TheInternationalConventionfortheSafetyofLifeatSea(SOLAS)

ThemainconventionformaritimesafetyistheSOLASConvention,adoptedinitsfirstversionalready

in1914.Theconventioncoversaverywiderangeofmatters,itsannexcontainingthesubstantive

rulesconsistsoffourteendifferentchapters.SomeoftherulesofSOLASareonlyapplicabletoshipsof

aspecifictypeoragewhiletheapplicabilityofothersdependsonthetradingarea.Thefocushereison

rulesapplicabletoanewbulkcarrierabove500gtincommercialuseininternationaltrade,witha

particularemphasisontherulesthatmayturnouttobechallengingforashipwithoutacrewon

board.Thebrief–andincomplete–reviewthusfocusesonoperationalandfunctionalrequirements

thatexplicitlyorimplicitlypresupposethepresenceofcrewmembers.

ChapterIestablishesthegeneralapplicationoftheregulationsintheAnnexandanexemptions

scheme,whichisbasedonthreedifferentcategoriesofexemptions:

1. CertaincategoriesofshipsthatarecompletelyexceptedfromtheSOLASrulesandhence

beyonditsscopearelistedinRegulation3.However,noneofthelistedcategoriesarerelevant

forpresentpurposes.

2. Regulation4(b)includesapossibilityforflagstateadministrationtoexempt“anyshipwhich

embodiesfeaturesofanovelkind”fromtherequirementsofChaptersII-1,II-2,IIIandIVif

theirapplication“mightseriouslyimpederesearchintothedevelopmentofsuchfeatures.Such

exemptionshallbecommunicatedtoIMOanddonotrelievetheshipfromtheobligationto

complywithsafetyrequirementsthatintheopinionoftheadministrationareadequateforthe

serviceandacceptabletothe(port)statestobevisitedbytheship.

3. Administrationshaveamoregeneralpossibilitytoacceptequivalentsolutionsiftheyare

satisfiedthattheequivalentisatleastaseffectiveasthatrequiredbytheConvention.More



42

specifically,thispossibilityapplieswhereSOLASrequires“thataparticularfitting,material,

applianceorapparatus,ortypethereof,shallbefittedorcarriedinaship,orthatany

particularprovisionshallbemade”.Inthesecasestheadministrationmayallowother

solutions“ifitissatisfiedbytrialthereoforotherwisethatsuchfitting,material,applianceor

apparatus,ortypethereof,orprovision,isatleastaseffectiveasthatrequiredbythepresent

regulations.”SuchequivalentsshallbecommunicatedtoIMOtogetherwithareportofany

trialsmade.

ChaptersII-1,II-2andIIIcontainrequirementsforshipsintheareasofstructure,stability,machinery

andelectricalinstallations(ChapterII-1),fireprotection(ChapterII-2)andlife-savingappliances

(ChapterIII).Thesechaptersmainlycoverconstruction,equipmentandmaterialsonboard,which

doesnotraiseparticularissuesfromtheperspectiveofautomatedoperations.Ashipthathastobe

constructedtomeetcertainstabilityrequirementsorfeaturessuchasdoublebottomswillobviously

havetodosoeveniftheshipisunmanned,andtheunmannedconditiondoesnotcallforadditional

requirementsinthisregard.

However,theyallincludesomedegreeofoperationalrequirements,relatingtoinformation

proceduresandcommunicationforthecrew,alarms,monitoringmechanismsetc.,whichare

obviouslydifficulttoapplyonacompletelyunmannedships.Insomecasesalarms,monitoring

equipmentandsystemoperationmayhavetobeshiftedoraddedtotheplacewherethecontrolleris

located,asotherwisethewholepurposeoftherequirementwouldbedefeated.Similarly,term

‘navigatingbridge’,whichfeaturesfrequentlyintherulesrelatingtosteeringgear,indicatorsand

varioustypesofengineandfirealarms,needtobeunderstoodasreferringtotheplacefromwhichthe

shipiscontrolled,iftherulesaretoretaintheirmeaningforremotelycontrolledships.Manyofthe

provisionsspecificallyaddressthepossibilitytoreplaceofhumanmonitoringbytechnicalequipment,

suchasunmannedmachineryspaces.

Forthesechapters,thepossibilityforexemptionsandalternativedesignsislikelytoplayanimportant

roleinfacilitatingcomplianceforunmannedships,2whichpresupposesthattheship’sflagstate

administrationisfavourabletoacceptingsuchexemptions.

TherequirementsconcerningradiocommunicationsinChapterIVincludefunctionalrequirements

ontheequipmentaswellaswatch-keepingrequirementsforthecrew.Thebasicfunctional

2 In addition to the exemptions provided for in Chapter I, the flag state administration may also under these three chapters exempt individual ships or classes of ships which do not proceed more than 20 nautical miles from the nearest land from the requirements “if it considers that the sheltered nature and conditions of the voyage are such as to render the application of any specific requirements … unreasonable or unnecessary” (Regulations II-1/1.4, II-2/4.1 and III/2.1).



43

requirementsarethatashipatseashallbecapableoftransmittingadistressalertbyatleasttwo

separateindependentmeans,receivingdistressalerts,communicating(transmittingandreceiving)in

distresssituations(searchandrescue),maritimesafetyinformation,generalradiocommunicationand

bridge-to-bridgecommunication.Whileatsea,everyshipshallmaintainacontinuouswatchona

numberoffrequenciesasprovidedfor.Therulesincludelimitedpossibilitiesforexemptionsand

alternativedesigns.Compliancewiththeserulesbyunmannedshipspresupposesthatradio

communicationcanberelayedtoaplacewhereacontrollerwithfullknowledgeoftheship’s

whereaboutsisoncall.

ChapterVcomprisesaverywiderangeofdifferentregulations,someofwhichmaybequite

challengingtoimplementforunmannedships,suchastherulesonmanningofships(Reg.14),voyage

planning(Reg.34),bridgevisibilityrequirements(Reg.22)orpilottransferarrangements(Reg.23).

Thechapteralsoincludesageneralobligationformasterstoproceedtotheassistanceofthosein

distress(Reg.33)andhighlightsthemaster’sdiscretionindecision-makingrelatingtosafetyofseaor

environmentalprotection,nottoberestrictedbytheowner,chartereroroperatingcompany(Reg.34-

1).ManyoftherulesofChapterVhavewiderapplicability,intermsofthesizeofshipsandtrading

areas,thantheotherSOLASchapters.Thescopeforexemptionsandequivalencesvariesfromone

regulationtoanother,butisingeneraltermsmorelimitedthangenerallyinSOLAS.

Therulesonmanningof shipsareofparticularrelevance.Generallyspeaking,decisionsonships’

manningarelefttotheflagstateadministration.Oncetheadministrationissatisfiedthatthenumber

andqualificationsofthecrewisadequatefortheshipinquestion,usuallyassessedonthebasisofan

estimateandjustificationproposedbytheship’sowner/operator,itwillissueasafemanning

documentfortheship.IntermsofsubstanceSOLASRegulationV/14essentiallyonlyrequiresthat

“fromthepointofviewofthesafetyoflifeatsea,allshipsshallbesufficientlyandefficientlymanned.”

Theassociatedguidelines(IMOResolutionA.1047(27))aremoredetailedandmentionabroader

rangeofobjectiveswithmanning,includingshipsecurity,safetyofcargoandenvironmental

protection,buttheyarenotlegallybinding.

Thekeyquestionwithrespecttounmannedshipsiswhethertheon-boardmanningcouldbereduced

totheextentthatasafemanningdocumentcouldbeissuedevenifthereisnotasinglecrewmember

onboardtheship,i.e.thatthesafemanningwouldbezero.This,inturn,iscloselylinkedtothe

questionofwhethertasksperformedbythecrewcanbetakenoverbyon-shorecontrollersor,inthe

caseofhighlyautomatedoperations,byotherpartiesresponsiblefortheship’soperations.



44

Ontheonehand,ifanationaladministrationweretodecidethatthefunctionsrequiredtoensurethe

safetyofoperationscouldbeperformedfromotherplacesthanfromtheshipitself,itisdifficulttofind

aprovisionthatwouldbedirectlyviolatedbythatdecision.‘Manned’isnotnecessarilythesameas

‘attended’andland-basedcontrollersofshipsmightverywellbeabletoperformmanyofthe

operationalfunctionsremotelywhileshore-basedmaintenancestaffcouldundertaketherequired

maintenanceandservicework.Indeedtheguidelinesonsafemanningspecificallyprovidethat

technicalequipmentandlevelofautomationistobetakenintoconsiderationwhendecidingonthe

manninglevels(Annex2,paras.1.1.3and1.1.4).Norwouldsuchadecisionnecessarilybeagainstthe

purposeunderlyingthesafemanningrules.Itisnotexcludedthattheoperationoftheshipmight

actuallygetsaferifmorefunctionsaretransferredtoshore,asnewtypesofequipment,redundancy

systemsetc.arebroughtonboardandnewfunctionswillbeperformedfromashore.

Ontheotherhand,theprecisewordingoftheindividualprovisionsshouldbeconsideredwithsome

cautioninthiscontext,asitisevidentthattheinternationalandnationalrulesonsafemanningare

draftedontheunderstandingthatthecrewisbasedonboardtheship.Theprospectofunmanned

shipswasnotthereatthetimetherulesweredevelopedandoneshouldthereforeavoidreadingin

toomuchsupportforthatdevelopmentintoexistinglegaltexts.Thisisallthemoretrueforfully

autonomousoperations,whichstretchesthenotionofmanningevenfurther.

ChapterVImainlycontainsoperationalrequirementsrelatedtothesafeloadingandunloadingof

solidbulkcargoes.Thechapter,likeEUDirective2001/96/EC,whichmakestheapplicationofthe

‘BLUCode’(IMORes.862(20))mandatoryinEUports,includesanumberofloadingproceduresand

requirementswhichpresupposeactivecommunicationbetweenthemaster,theshipperandthe

terminaloperator.

ChapterIXmakesmandatorytheInternationalSafetyManagement(ISM)Code,whichrequiresa

safetymanagementsystemtobeestablishedbytheshipowneroranypersonwhohasassumed

responsibilityfortheship(the"Company").ThemainpurposeoftheISMCodeistoachieveagreater

involvementoftheshore-sidecompanyinthesafetymanagementofindividualships.Itincludes

requirementsondefiningthemaster’sresponsibilities,plansforshipboardoperationsand

maintenance,emergencypreparedness,documentationetc.

Evenifunmannedshipoperationswillinevitablyservetostrengthenthelinkbetweenshore-based

operatorsandtheship,compliancewiththeCodeposescertainchallengesincasethemanningofa

shipconcernedisreducedtozero.Thisisparticularlythecasewithrespecttolinesofcommunication

andreportingrequirements.ItcanbefurthernotedthatSOLASincludesnopossibilitiesfor



45

exemptionsfromChapterIX,exceptforgovernment-operatedshipsusedfornon-commercial

purposes.

ChapterXI-2addressesmeasurestoenhancemaritimesecurity.Itmostlydealswithobligationsfor

(flagstate)administrationsandshipoperatingcompanies,butpresupposeaclosecommunication

betweenthemandtheship.Regulations11and12specificallyprovideforthepossibilityforstates

partiestoagreeonalternativesecurityagreementswithotherstatesorequivalentarrangementsfor

theirownshipsprovidedtheyareatleastaseffectiveasthoseprescribedinChapterXI-2.

3.3InternationalConventionforthePreventionofPollutionfromShips(MARPOL)

MARPOListhemainIMOconventionfordealingwithvariousformsofpollutionfromships.Itincludes

constructionandequipmentprovisions,e.g.foroiltankers,butalsocertainoperationaland

proceduralrequirements,includingdischargelimits,proceduresforship-to-shiptransfers,various

reportingobligationsincaseofspillsandrequirementstokeepdifferentrecordbooks.Theapplicable

requirementswillnodoubthavetobecompliedwithbyunmannedships,butgenerallyspeakingthe

MARPOLrequirementsareunlikelytopresentparticularchallengesinthisregard.Recordbookscan

presumablybemaintainedinanelectronicformatwhilereportingandnotificationobligationsexistin

severalconventionsandneedtobeaddressedinsimilarways.Responsestopollutionemergenciesas

outlinedinthe‘shipboardoilpollutionemergencyplan’(SOPEP)willhavetobeadaptedtothe

responsecapabilitiesofunmannedships.

3.4ConventionontheInternationalRegulationsforPreventingCollisionsatSea(COLREGs)

TheCOLREGsincludeavarietyof‘rulesfortheroad’inshipping,includingonsafespeed,signals,

lights,etc.andrulesonprioritiesandmanoeuvringfordifferenttypesofvesselsindifferentsituations.

Theruleswillalsoapplytoanunmannedship,whichrepresentnospecialcategoryofshipswithinthe

meaningoftheCOLREGs.

TheCOLREGscoverbothcorenavigationaltasksofthecrewonboardaship:situationawareness

(includinglookout)andoperationaldecision-makingwhenitcomestocollisionavoidance,priorities,

speedetc.Bothaspectsarelikelytoposechallengesforunmannedships.

Thelook-outrequirementisprovidedforinRule5:



46

“Everyvesselshallatalltimesmaintainaproperlook-outbysightandhearingaswellasbyallavailablemeans

appropriateintheprevailingcircumstancesandconditionssoastomakeafullappraisalofthesituationandof

theriskofcollision.”

Thepurposeofthelookoutruleistomakesurethatwhoevercontrolstheshipareawareofthethings

aroundthemtomakeinformeddecisionswithrespecttoactionsinavoidingcollisions.Thetermlook-

out,asusedbytheRules,doesnotnecessarilydenoteaperson,butratherthesystematiccollectionof

information.Moreover,theuseofvaguetermssuchas“proper”and“appropriate”providesflexibility

forhowsuchlook-outisorganisedonboard.3

ThekeyquestionforunmannedshipsiswhetherthewordingofRule5isbroadenoughtoauthorisea

replacementofthehumanlookoutbyvarioustypesofcameras,radar,audiotechnologyandother

technicalsolutions.Onthebasisofthepurposeoftheruleanditsflexiblewording,itisarguablethat

thiscouldbeacceptediftheequipmentallowsthecontrollertohaveanadequateoverviewofthe

circumstancesallowinghimtakeappropriateactioningoodtime,tothesameextentorbetterthanif

hewouldbeonboard.However,inviewofthewidespreadauthorityofCOLREGsandthenatureof

collisionregulation(alwaysinvolvingmorethanoneship),anysuchclarificationorinterpretation

shouldbedoneatinternationallevelratherthanbyindividualstates.

Aseparatequestioniswhethertheremotecontrollercouldalsobeinchargeoftherelevant

operational decisionsontheship’snavigationandmanoeuvring.Forthismatter,COLREGsdonot

poseanydirecttextualobstacle.Thesubjectsofthesteeringandsailingrulesare‘vessels’,withoutany

furtherdetailsaboutthepersonbehindthedecisions.Themoreproblematicquestionariseswhen

operationaldecisionsareautomated,withoutacontrollerinchargeofthecompletedecision-making.

Fromatechnicalpointofviewitisprobablyfeasibletocreatealgorithmsthatcomplyverydiligently

withthesteeringandsailingrulesofCOLREGs,eventakingintoaccountthesometimesunpredictable

actionsofotherships.Achallenge,however,isthattheCOLREGsdonotofferabsoluterulesof

conduct.Therulesforpreventingcollisionsincludeobligationforbothvesselstotakeavoidance

actionifitseemsthatthereisariskofcollision.Inaddition,theCOLREGsincludearulewhichserves

togiveprecedencetogoodseamanshipoveritsownprovisions.4Whatconstitutesgoodseamanship

3 See e.g. C. Llana & G. Wisneskey Handbook of the Nautical Rules of the Road, 3rd on-line edition, 2006 (updated in 2011), available at http://navruleshandbook.com/Rule5.html 4 COLREGs Rule 2 provides that:

(a) Nothing in these Rules shall exonerate any vessel, or the owner, master or crew thereof, from the consequences of any neglect to comply with these Rules or of the neglect of any precaution which may be required by the ordinary practice of seamen, or by the special circumstances of the case. (b) In construing and complying with these Rules due regard shall be had to all dangers of navigation and collision and to any special circumstances, including the limitations of the vessels involved, which may make a departure from these Rules necessary to avoid immediate danger.



47

forthispurposeisamatteroffacttobeassessedafterconsiderationofallrelevantprevailing

circumstances.Itseemsclearthattheincorporationof‘goodseamanship’intoanyautomated

navigationprogrammemaybecoupledwithseriousdifficulties.

AnotherquestionlinkedtoCOLREGsiswhetherunmannedshipsshouldbegivenaspecificsignal,

light,AISmessageortheliketoinformmarinersonboardothershipsabouttheirstatus.Theanswer

tothisisprobablypositive,andalthoughsomenationalsolutionsinthisfieldcouldbejustifiedunder

Rule1(b),suchdecisionsshouldpreferablybemadeatinternationallevel.Bycontrast,iftheobjective

isthatunmannedshipsshouldbenaturallyintegratedintotheenvironmentofmannedships,itdoes

notseemjustifiabletomaintainthatunmannedshipsshouldhaveastatusthatwouldofferitspecial

privilegesandprioritiesoverothershipsinCOLREGs.5

3.5InternationalConventiononStandardsofTraining,CertificationandWatchkeepingforSeafarers

(STCW)

TheSTCWConventiondoesnotstrictlyspeakingapplytopersonswhoarenotworkingonboard

ships.AccordingtoitsArticleIII,theConventionapplies“toseafarersservingonboardseagoingships”

flyingtheflagofastateparty.

Evenifnotstrictlyspeakingapplicable,itisevidentthatacorrespondingtrainingregimewill

eventuallyhavetobedevelopedforpersonsoperatingshipsremotely.Intheshorterterm,national

administrationshavebeengrantedsomediscretiontoapplyequivalentarrangements,includingto

caterfortechnicaldevelopments.UnderArticleIX(1):

TheConventionshallnotpreventanAdministrationfromretainingoradoptingothereducational

andtrainingarrangements,includingthoseinvolvingseagoingserviceandshipboard

organisationespeciallyadaptedtotechnicaldevelopmentsandtospecialtypesofshipsand

trades,providedthatthelevelofseagoingservice,knowledgeandefficiencyasregards

navigationalandtechnicalhandlingofshipandcargoensuresadegreeofsafetyatseaandhasa

preventiveeffectasregardspollutionatleastequivalenttotherequirementsoftheConvention.

Thequalificationandcompetencesofpersonnelwhoareoperatingshipsfromaremotelocationneed

tobegivenconsiderationinviewofthecombinationofmaritimeandtechnologyskillsthatisneeded

forthistypeofwork.Inthemeantime,itisprobablysafetoapply(atleast)theSTCWandother

5 It has been suggested, for example, that compliance with the COLREGs might be ensured merely by treating unmanned ships as a vessel ”not under command” or ”restricted in her ability to manoeuvre” under Rule 18, which would require other ships to give way.



48

nationalrequirementsanalogically(asifthepersonswereonboardtheship).Ifandwhenitis

consideredthatunmannedshipoperationsrequireparticulartraining,therelevantprovisionswould

probablyneedtobeamendedtoaccommodatethenewrequirementsfortheoperationofunmanned

orlargelyautomatedships.

ThebiggestchallengesforunmannedshipsinrelationtoSTCWprobablylieinthefieldof

watchkeeping.Theresponsibilitiesforsafewatchkeepinginvolveseveralpersons,includingthe

company,themaster,thechiefengineerofficersandthewholewatchkeepingpersonnel,whose

responsibilityitistoensure“thatasafecontinuouswatchorwatchesappropriatetotheprevailing

andconditionsaremaintainedonallseagoingshipsatalltimes”.This,accordingtoRegulation

VIII/2(2)(1),includesthat“officersinchargeofthenavigationalwatchareresponsiblefornavigating

theshipsafelyduringtheirperiodsofduty,whentheyshallbephysicallypresentonthenavigating

bridgeorinadirectlyassociatedlocationsuchasthechartroomorbridgecontrolroomatalltimes.”

ThemoredetailedrequirementsarelaiddownintheSTCWCode,whichinitsmandatoryPartA

includesdetailedprovisionsforwatchkeepinginvariousconditions,includingrequirementson

lookout,bridge,engineroomandradiowatches.Provisionsforworkhoursandrestinghoursare

includedintheactaswellasanobligationtoperformrouteplanningaheadoftheintendedvoyage.

Itisprobablydifficultforunmannedshipstomeetthewatchkeepingrequirementaslaiddowninthe

STCWConventionandCode,whichsuggeststhatsomeamendmentoftheseinstrumentswillbe

necessarybeforecommercialshipscanoperatecompletelywithoutacreworevenwithradically

reducedwatcharrangementson-board.Ontheotherhand,itshouldbeborneinmindthatthe

reductionofon-boardcrewwillnormallybecompensatedbyotherfunctionsperformedremotely.

Theseland-basedfunctionsshouldatleasttosomeextentalleviatetheconcernsrelatedtofatigueand

reductionofsafetylevelswhichareusuallyassociatedwithreductionsofon-boardcrew.

Intheend,thedecisionofwhetheraparticularmanningsufficesformaintainingasafelookoutand

watchkeepingontheshipwillhavetobeaddressedthroughtheprocessofsafemanningwhereall

suchfactorswillhavetobetakenintoaccount.

3.6MaritimeLabourConvention(MLC)

Theprincipalconventioninthefieldofmaritimeemployment,the2006MaritimeLabourConvention

(MLC)addressesarangeofissuesrelatingtolabourconditionsonboardships,rangingfrom



49

recruitmentandconditionsofemploymenttofundamentalrightsofseafarersandrecreational

facilitiesonboard.Itappliestoallseafarersonships“ordinarilyengagedincommercialactivities”.

ThefirstpointtobenotedwithrespecttounmannedshipsisthatthescopeoftheMLCConventionis

limitedto‘seafarers’(MLC,ArticleII(2)),whichisdefinedinMLCArticleII(1)(f)as“anypersonwhois

employedorengagedorworksinanycapacityonboardashiptowhichthisConventionapplies”

(emphasisadded).Literallyspeaking,ashipwhichisentirelyunmannedisaccordinglynotsubjectto

theserules.Yet,forunmannedshipsitmightnotbethelastword,asArticleII(3)includesaspecific

procedureforsettlingwhetheraparticularcategoryofpersonistoberegardedasaseafarer:

“Intheeventofdoubtastowhetheranycategoriesofpersonsaretoberegardedasseafarersfor

thepurposeofthisConvention,thequestionshallbedeterminedbythecompetentauthorityin

eachMemberafterconsultationwiththeshipowners’andseafarers’organizationsconcerned

withthisquestion.”

Secondly,sincetherulesmainlytargetlivingandworkingconditionsonboardshipstheircontent

largelylosetheirrelevanceiftheshipiscompletelyunmanned.Itisthereforelikelythatissuessuchas

employmentconditions,workinghoursetc.forshore-basedremotecontrollerswillbesubjectto

relevantland-basedrules,possiblytobecomplementedbyseparateruleswhichtakeinto

considerationsthespecificnatureoftheirtasks.

4.Liabilityrules4.1General–AutonomousSystemsChallengeLegalThinking

Inviewoftheprojectedincreaseofautonomousvehicletechnologies,futureaccidentswill

increasinglybecausedbydefectiveproductsandsystems,whiletheroleofhumanerrorisreducedor

atleastshiftedelsewhere.Whenthereislesshumancontrol,thereliabilityandproblem-solving

capacityofanautonomoussystembecomecrucial.Theautonomoussystemmustsurviveevenwhen

humaninterventionisnotpossible.Thisalsomeansachangeinlegalthinking.Liabilityfordamages

cannotbebasedonhumanactsoromissionsinthesamewayastoday.

Currentlylawdoesnotprovideclear-cutanswerstoquestionsonliabilityforautonomousoperations.

Intheory,severalactorsmaybeheldliableforaccidentscausedbyanautonomoussystem.Liability

could,forexample,beplacedwiththeowner,userormanufacturerofanautonomousdevice,oreven

onthemanufacturerwhohasproducedthedefectivecomponent.Asautonomoussystemsbecome



50

morecommon,thequestionofliabilityneedstobeclarified.Somemanufacturersofself-drivingcars

haveevenvoluntarilytakenupthequestion,irrespectiveofthelegalframeworkinvolved:

“Wearethesuppliersofthistechnologyandweareliableforeverythingthecarisdoingin

autonomousmode.Ifyouarenotreadytomakesuchastatement,youshouldn'ttrytodevelopan

autonomoussystem.”6

Fromatechnologypointofview,autonomousvesselsandself-drivingcarsmayhavemanythingsin

common.Nevertheless,legalconclusionsappliedtoroadtrafficarenotdirectlytransferableto

autonomousshippingandviceversa.Actors,automation,accidentsandcontextaredifferent.For

example,shipsaremorelikelytobeoperatedbycompaniesthanbyprivateindividualsand

automationismorefocusedonremotecontrolthancompleteautomation,atleastintheearlyphases.

Inthefollowing,thebasisofthecurrentmaritimeliabilitylegalframeworkispresentedinsection4.2.

Afterthat,somewaysinwhichautonomoustechnologiesmayaffectthefunctioningoftheexisting

liabilityframeworkarehighlightedinsection4.3,followedbysomeconcludingobservationsinsection

4.4.Asliabilityregimesdifferindifferentjurisdictions,thepresentoutlinespecificallydepartsfrom

theNordicand,inparticular,Finnishlegalperspective.

4.2MaritimeLiabilityRules

Maritimelawrelatingtoshipoperators’liabilitiesandcompensationofdamageincludeanumberof

peculiaritiesthatarespecificforthisbranchoflaw.Theruleshavebeendevelopedwiththeparticular

featuresofshippinginmind,oftenoriginatinginconsiderationsandconceptsthathavebeenapplied

forcenturies.Basicissues,suchaswhoisresponsible,onwhatbasis,andforwhatamounthaveto

someextentbeenharmonisedthroughinternationalconventions.However,significantnational

variationsexistasstates’participationtothemaritimeliabilityconventionsisnotasuniformasforthe

safetyconventionsdiscussedaboveandasliabilityissuestoalargerextentdependonnational

traditionsandthelegalsystemconcerned.Whatlawswillbeappliedinagivencaseinturndepends

onaseriesoffactors,includingwheretheincidenttookplace,thetypeofincidentsand,insomecases,

onthenationalityofthekeyplayersinvolved,includingtheship’sflagstate.

First,withregardtotheliableperson,existingmaritimeliabilityrulesgenerallychannelsliabilityand

dutiestoowners/operatorsofships(Finnish:‘laivanisäntä’,German:‘Reder’,French:‘armateur’),

6 Håkan Samuelsson, President and CEO of Volvo Car Corporation. See www.autoblog.com/2015/10/09/volvo-accept-autonomous-car-liability/



51

ratherthantoindividualcrewmembersorotherassistants.Likeemployersgenerally,the

owner/operatorhasabroadvicariousliabilityfordamagecausedintheservicebythefaultorneglect

ofthemaster,crew,pilotorothersperformingworkintheserviceoftheship.Thepossibilityfor

aggrievedpartiestoclaimdamagesfromothersthantheowner/operatorislimited,butthe

owner/operatorhimselfmaybeabletotakesubsequentrecourseactionagainstthepartyatfault.

Specialliabilityrulesforparticularcasesmayalterthisstartingpoint,butthebroadvicariousliability

oftheowner/operatorremains.Incaseofliabilityforcollisions,forexample,liabilityisplacedonthe

‘ship(s)’atfaultwithoutanymentionofthepersonsactuallybehindthecollision.Environmental

liabilityrulesnormallychannelliabilityexclusivelytotheregistershipowner,specificallyexcludinga

rangeofotherpotentiallyliablepersons.Theidentityofthepersonwhosefaultactuallycausedthe

damagewillthereforenotnormallyhaveanimpactonthequestionofliablepersonfromthepointof

viewofclaimants,aslongasthefaultissomehowlinkedtotheoperationoftheship.

Second,astothethresholdoffaultornegligencerequiredtotriggerliability,therulesdifferbetween

differenttypesofliability.Forcertaincases,suchasincidentscausingpollutionorinjuryto

passengers,it’sacceptedthatclaimantsneednotdemonstratenegligenceonbehalfofthe

owner/operatortobecompensated(i.e.ownersinthesecaseshaveastatutory‘strict’liability).Inthe

absenceofsuchrules,thegeneralruleisthatliabilityoftheowner/operatorpresupposesfault

(negligentactsoromissions)onbehalfoftheowner/operatororhishelpers.Fault-basedliabilityis

alsothesoleruleforapportioningliabilityincaseofcollisions.Autonomousshipoperationsmay

introducenewconsiderationsregardingfaultwhicharediscussedinsection4.3below.

Third,currentmaritimelawgrantstheliablepartywithawide-reachingrighttofinanciallylimitthe

liabilityperincidentbasedonthesizeoftheship.Therightofshipowners/operatorstolimitliability

islostonlyinveryexceptionalcases.Claimantsmayaccordinglynotbeabletorecoverfull

compensationfortheirlosses,howeverlegitimatetheirclaimsmaybe.Limitationofliabilityappliesto

faultscommittedbypersonsforwhomtheshipowner/operatorisresponsibleandhencetoabroad

numberofhelpersinvolvedintheoperationoftheship.

Thekeyelementsofthegeneralmaritimeliabilityregimearethusabroadvicariousliabilityplacedon

theowner/operatoroftheship,whichisbasedonfaultorneglectandprotectedbyastrongrightof

limitation.Theserulesalsoformthebasisforliabilityinsurancesandotherriskmanagement.For

shipsabove300gtenteringEuropeanUnionportsthereisanobligationtomaintainliabilityinsurance

uptotheapplicablefinanciallimits.

4.3ImplicationsofAutonomousShipping



52

Eveniftheremaynotbeanimmediateneedtochangethefoundationsofmaritimeliabilityfor

autonomousships,itisneverthelessimportanttorecognisethatthetechnicaldevelopmenttowards

increasedautomationdoesinvolvecertainchallengestothecurrentliabilityframework.

Whileerrorscommittedbypersonscontrollingremotely-operatedshipsareprobablytobetreatedin

thesamewayaserrorscommittedbyon-boardcrewmembers,autonomoustechnologymaygenerate

newtypesoferrorsandcausalrelationships.Oneexampleisdamagecausedbymalfunctionofan

autonomoussystem,e.g.bydevicefailureorfaultysoftware.Evenunderthesecircumstancesthe

owner/operatorwouldprobablybeliable,atleastinpart,ifhe(orhisassistants)failtooverridethe

autonomoussystem.However,scenarioswherehumaninterventionisnotevenpossiblearemore

complicated.Forexample,iftheconnectionbetweenthevesselandcontrolleriscutoff,thevesselwill

havetorelyexclusivelyonitsautonomoussystems.Ifanaccidentthenoccursduetofailuresinthe

autonomoussystem,duetowrongfulprogrammingetc.,itislessobviousthattheowner/operator

wouldcarrytheliabilityunderastrictlyfault-basedliabilityscheme.

Suchdrawbacksofafault-basedliabilityschemeforhighlyautomatedsystemsmayadvancethe

argumentinfavourofastrictliabilityregimeforautomatedships.That,ontheotherhand,would

createasignificantdifferentiationbetweenmannedandunmannedvesselswhichmightnotbe

justifiedfromariskpointofviewandwouldinanycaseresultindifficultissuesofdelimitationand

definition.

Asanalternative,claimantsmaytrytobasetheirclaimsonotherliabilitysystemsthanthemaritime

one.Ifaccidentswereincreasinglycausedbydefectiveautonomoussystems,theaggrievedparties

couldtrytomakeclaimsagainstthebuilderofthevesselorthemanufactureroftheautonomous

system,itssoftwareetc.Thiswouldmeanashifttowardsproductliabilityinthemaritimecontextto

fillaperceived‘liabilitygap’inmaritimelaw.Thedevelopmentcouldbeadvantageousforclaimants,

as,forexample,theEUdirectiveonproductliabilityisbasedonastrictliabilityoftheproducerand

doesnotincludeageneralfinanciallimitationofliability.

Itseemsinevitablethatpressuresforsuchalternativesolutionswillgrowifitturnsoutthatthe

existingmaritimeliabilityregimeisinsufficienttocovertheconcernsofbusinesspartners,claimants

andthegeneralpublicrelatingtotherisksinvolvedwithautonomousshipping.Autonomousshipping

mayverywellactasacatalystforthisdevelopment,asitiseasiertoappreciatethecriticalroleofthe

product(liability)insystemswherethereisnohumaninterventioninvolved.



53

Figure1:Thepossibleliabilityframeworkinautonomousshipping

Itshouldbeemphasised,however,thattheapplicationofproductliabilityrulestoautonomous

shippingisbynomeansstraightforwardeither.TheEUDirectiveonthematter,forexample,only

coversalimitedrangeofthepotentiallyrelevanttypesofdamages.Forafullerpictureofroleof

shipyardsandmanufacturersofautonomoussystems,othersupplementaryliabilitysystemsmustalso

bestudiedinparallel.Thekeypointatthisstageismerelythatproductandotherliabilityrulesmay

verywelloperateinparallelwiththetraditionalmaritimeliabilityregimeinthefutureandthatthe

prospectofseveralbasesofliabilityforautonomousshippingneedstobetakenseriouslyfromthe

outset.Thepresenceofparallelliabilityregimesnecessarilyinvolvescomplexlegalquestionsrelating

toscopeandpriorities.

4.4ConcludingObservations

Autonomousshippingmightnotimposeasacutedemandsforchangeofthemaritimeliabilityrulesas

isthecaseforsomeoftheIMOConventionsdiscussedinsection3.However,itwillaffectthemaritime

liabilityframework,possiblyquitesignificantly,albeitataslowerpace,initiallyprobablydrivenby

nationalcaselaw.Inthelongerterm,however,autonomousshippingcouldcontributetothe

introductionofnewlegalregimestosupplementthetraditionalmaritimelawframeworktofill

(perceivedorreal)gapsintheexistingmaritimelawregime.Astherateofautomationincreases,there

needstobetrustnotonlyinthesystemsassuch,butalsointhelegalregimewhichistheretomake

goodforanydamagecausedbythenewtypeofoperations.

Theincreasedautomationinshippingmayalsoaffectmaritimeriskmanagementmoregenerally.

Currentinsuranceandcontractualarrangements,forexample,areallbasedonthepremisethatships

aremanned.Inautonomousshipping,theplayersinvolved,theirroles,responsibilitiesandliabilities

willbedifferent,whichcallsforconsequentialadjustmentsininsuranceandcontractualpractices.The

legalimplicationsofautonomousshippingaccordinglyextendbeyondtheliabilityrules.



54

5.Summary

Theexistingmaritimelawframeworkdoesnotanticipateunmannedshipping.Abroadrangeofrules

arepotentiallyconcernedbyashifttounmannedshippingoperations,butthenatureofthechallenge

toaccommodatethisshiftintheexistinglawdiffersfromonetypeofruletoanother.

Sinceitisassumedthatthevehiclesofinterestherequalifyas‘ships’underthevariousinternational

andnationalrules,theregulatorysituationisreasonablystraightforward.Thestartingpointisthatthe

unmannedshipsaresubjecttothesamerightsandobligationsastheirmannedcounterparts.

Themostimmediatechallengesforensuringthelegalityofunmannedshippingoperationsarefound

atthelevelofinternationaltechnicalrules,i.e.theIMOrules.Thisisnotonlywherethemostclear

substantivetensionsarefoundinrelationtoexistingrules,buttheserulesarealsodecisivefor

steeringthecontentofthejurisdictionalrulesofthelawoftheseaaswellasofnationalmaritimelaws

worldwide.Inotherwords,ifIMOrulesspecificallyrecognisedandauthorisedunmannedshipping

operations,evenasanoption,theregulatorychallengeattheotherlegallevelswouldbesignificantly

reduced.

Thenatureofthechallengealsodependsonthetradingareaoftheshipandofthelevelofautomation

concerned.Shipmovementswithinasinglestateinvolvesonestate’sapprovalonlyandalargepartof

theinternationalrequirementsdonotapplytosuchtransports.Autonomousshipsinvolvegreater

legalchallengesthanremotelyoperatedones.Thelatteronesstillhaveacrew,evenifnotonboard,

andmayhencemoreeasilycomplywithanumberoftoday’soperationalrequirements.

TheIMOrulesinternationalrulesacceptasignificantdiscretionfortheflagstateadministrationto

acceptalternativeandequivalentsolutions,whichwillnodoubtbeofrelevanceintheearlyphasesof

unmannedshipping.ThisflexibilityforflagstateshasbeensomewhatreducedforEUmemberstates

bytheintroductionofEUmaritimelegislation,butitisclearthatunmannedshippingcannotbe

introducedintheearlyphaseswithoutsignificantco-operationbytheship’sflagstateadministration.

Maritimeliabilityrulesseemlessacutetoamend,butarealsolikelytoundergosignificantchanges

overtime,asnewplayers,newrisksand-possibly-newliabilitysystemswillenterthescenewith

unmannedshippingoperations.Existingliabilityrulesmayneedtobeinterpreted,amendedand

possiblysupplementedbydedicatedrulestosupplementthetraditionalmaritimeliabilityframework.



55

Newliabilityrules,inturn,willhaverepercussionsonmarineinsuranceandotherbusiness

relationshipsoftheshipoperators.

Thelegalchallengesdiscussedherearenotinsurmountableaslaws,atanylevel,canalwaysbe

amendedtoaccommodatenewdevelopments.Thebiggerquestioniswhetherthereissocietal

acceptanceandpreparednessinthemaritimecommunityandbeyondtomakechangesto

accommodateunmannedshipping.Iftheanswertothatquestionispositive,thelegalchallengeis

reducedtoidentifyingthekeyrulesthatareinneedofadjustmentsandmaketheamendments.The

amendmentscouldpossiblyevenbeintheformofagenericacceptanceofcertainkeyissuesof

principle,suchasthepossibilitytoperformon-boardfunctionsfromaremotelocationandthe

relationshipbetweencrewresponsibilitiesandautomatedfunctions.

Suchinternationalamendments,however,takeseveralyearstoinitiateandformulateandstillmore

yearstocomeintoeffect.Intheinterim,non-bindingIMOguidelinesorbestpracticecodesfor

unmannedshippingoperationsmayprovideimportantsupportandassistanceforflagstatesthatsee

thebenefitsofthedevelopmentandwishtosupportit,butarestillnotpreparedtoriskthe

internationalconnectionthathasinspiredmaritimeregulationforcenturies.

Annex:Summaryofthedifferentlayersandsubstantivebranchesofmaritimelaw

Jurisdictional

rules

(maintarget:states)

Technicalreq. and

standards

(maintarget:flagstates)

Privatelawissues

(shipownerandothercommercial

partners)

Otherrules

(Criminal,social,

commercial,publiclawetc.)

Global(UN) UNCLOS

Global

(IMO&ILO)

SOLAS,MARPOL,STCW,

COLREGS,MLC

Global(IMO,

UNCITRAL,

CMIetc.)

Privatelawconventionsone.g.

liability,limitation,arrest,carriage

ofgoods,salvage,etc.

European

Union

Shipsafetydirectives&

regulations

Limitationsonexemptions

Productliabilityrules,insurance

requirements

Rulesoncompetentjurisdiction

andapplicablelaw

SeveralissuescoveredbyEU

Treaty&legislation

Nordicstates NordicMaritimeCodes,Nordic

marineinsuranceterms

National

(Finland)

Nationalimplementing

legislation,discretionofflag

stateadministration(Trafi)

FinnishMaritimeCode674/1994,

otherspecifiedactsonliability,

insuranceetc.

Theentirelegislationapplies

aprioriforshipsflyingits

flag



56

SafetyandsecurityRistoJalonen,Projectmanager,seniorresearchscientist,AaltoUniversity

DepartmentofMechanicalEngineering/MarineTechnologyGroup

RistoTuominen,Seniorscientist,VTTTechnicalResearchCentreofFinlandLtd

MikaelWahlström,Seniorscientist,VTTTechnicalResearchCentreofFinland

Ltd



57

Safetyandsecurityinautonomousshipping–challengesforresearch

anddevelopment

SummarySafetyandsecurityimposeessentialconstrainingrequirementsthatneedtobefulfilledinthedesign

andimplementationofshipautomation.Inprinciple,autonomousortele-operatedshipsarerequired

tobe,atleast,assafeasconventionalvesselsinsimilarservice.However,duetoconsiderable

uncertaintyconcerningnewhazardsandrisks,itmaybepossiblethatevenmorestringentsafetygoals

areneededforfutureapplicationswithexpandedportionsoftasksandoperationscarriedouteither

underremotecontrolorasautonomousoperations.Problemscanbetreatedaschallenges,andin

engineeringtheymayoftenbesolvedbycreatingnewtechnicalorsometimeseventechnological

solutions.Thequestions,whatneedstobedonetoensuresafetyandsecurityinshipswithcontinually

risinglevelsofautomationandremotecontrol,oruptowhatleveltheautomationcanbeincreasedin

ships,havebecomemorerelevantthaneverbefore.Whileaddressedinitiallyinafewearlierstudies,

theimpactsofautonomous,unmannedmerchantshipsonmaritimesafetyhavenotbeenstudied

widelyanddeeplyenough,yet.ThegapsininformationwillbefilledtosomeextentintheAAWA

Initiative.

Makingsomethingnew–somethingthathasnotexistedbefore–iscentraltoengineering.According

tohistoricalrecordstheconceptoffailurehasquiteoftenbeencentraltotheincreasingunderstanding

inseveralareasofthemultidisciplinaryengineeringscience.Innewdesignsithasalwaysbeenan

overridingobjectivetoavoidfailure.Safetyandsecurityneedtobetakenintoaccountwellenough

fromthebeginningofthedesigntotheendofthewholelife-cycleofthenewdesign.Thesocio-

technicalapproachhaswidenedoureyestopossibilitiestoconsider,notonlymanyimportant

technicaldetails,butalsowideraspects.Thisleadsustosystemicthinkingwiththeimportanteffects

ofoperationalandorganisationalfactorsshapingthesystemdesign.

Visionsofnewdevelopmentsinthefieldofautomationmaybeidentifiedasheraldsofanew

technologicalandoperationalera.Emergenttechnologymayincludemanyhazardsandevensome

disruptiveeffects.Thesefeaturessethighdemandsonthesocialresponsibilityofdeveloperstocover

allimportantaspectsintheirassessmentsofimpactsonsafetyandsecurity.So,obligationsfor

meticulousandover-archingworkbeforepracticaltestapplicationsandfirstcommercialsolutionscan



58

starttospreadarehigh.Oneoftheproblemsrelatedtothemanychallengesistoassessthepaceofthe

development.Ifthetechnologicaldevelopmentseemstobefasterthanthedesignandconstructionof

allnecessaryandfeasiblesafetylevers,moreeffortsshouldbeputongettingthemdevelopedatthe

samepace.

Alongsidedevelopmentoftechnologicalsolutionstoenablehigherlevelsofautonomousoperationof

ships,theAAWAinitiativeaimstobuildupawarenessandunderstandingonsafetyandsecurityrisks

relevanttoenvisionedautonomousconcepts,andpointoutsomesuggestedmeasurestomanagethese

riskseffectively.Theriskknowledgewillbebuiltupgraduallyandcumulativelythrough

comprehensiveanalyses,simulatorstudies,andfinallyinpilotdemonstratorstudiesandteststobe

executedonactualseagoingvesselswithsomethoroughlyconsideredrestrictions.Understanding

autonomousortele-operatedshipsystemsandtheirembeddedcomplexitygrowsgradually,butitis

alsoimportanttoformaholisticpictureofthenew,emergenttechnologyunderdevelopment.Thisis

whatweaimatinAAWA.

1. IntroducingofautonomousmerchantshipsformaritimeoperationDesignandimplementationofmerchantshipconceptsplannedtooperatepartiallyorfully

autonomouslyorunderremotecontrolfromashore,isstillinitsinfancy.Howeverthisvisionofnew

eraofmarinetransportationisgainingincreasinginterestamongthemaritimeindustryandnew

conceptsaredevelopingquicklyworldwide.Theadvancesininformationandcommunications

technology(ICT)inrecentyearsenablequickdevelopmentastheymakepossibletheon-board

intelligenceanddataconnectioncapabilitiesnecessaryformakingshipsabletooperateevenwithout

on-boardcrew.

Theeconomicbenefitsofautonomousoperationconceptshavebeenhypothesisedtocapitalize

highestinoceangoingfreightvesselstransportingrelativelylowvaluecargosonintercontinental

routes.However,mostlikely,thefirstimplementationsincommercialtraffic,firstpilotingandthenin

operationaluse,couldbeexpectedinshortseatrafficandspecialtypeofapplicationswhichoperatein

nationalwaters.Thisisbecauseoftheassociatedeconomicrisk,andtheneedtohavehighconfidence

ontheperformance,reliabilityandsafetyofthesolutionsproposedbeforetakingthemintodeepsea.

Anotherincentivesuggestingthiskindofdevelopmentpathcomesfromtherestrictionsimposedby

themandatoryinternationalmaritimeregulationswhichdonotcurrentlyrecognizetheconceptof

unmannedshipoperation.Consequently,autonomousoperationswillinitiallyrequireexemption

permitswhichacompetentflagstateadministrationmayissueonaparticularshipfornationalwaters

ifsafetyandsecurityarenotcompromised.



59

Dependingonhumanoperators’presenceandinvolvementinmonitoring,planning,executionand

controlofshipoperations,differentlevelsofshipautonomycanbeidentified.However,muchworkon

standardizedclassificationshelpingtoidentifyandspecifythedifferentautonomylevelsinshipsmay

stillbeexpected.Inprinciple,inthelowerormediumlevelsofshipautonomy,theincreased

intelligenceintroducedon-boardwouldjustprovideextraassistancetothebridgeoperators,orit

wouldtakeoversmallerorlargerpartsofbridgeoperatortaskswhilebeingundersupervisorycontrol

byacompetentcrewmemberremainingpresenton-boardandabletointerveneincaseofproblems

identifiedbythesupervisororbytheautomatedsystem.

Inhigherlevelsofshipautonomy,thesupervisorycontrolpartmaybetransferredtoadedicated

ShoreControlCentre(SCC)whereasupervisormaybeabletomonitortheoperationofseveralvessels

simultaneouslyandinterveneremotelywhenaspecificneedisidentified.TheSCCcouldalsohavethe

responsibilityforexecutingspecificoperationsofaship,e.g.steeringinandoutaport,whichthen

wouldbecarriedoutremotelybytele-operation.Inthehighestlevelofshipautonomy,theships

wouldoperatewithoutcontinuoushumansupervisionpresenteitheron-boardnoratsomeonshore

controlcentre.However,theautomatedsystemisplannedtomakecontactwithanSCCforhelpwhen

encounteringaproblemsituationitisnotabletoresolve.Suchconnectivity,alwaysavailablewiththe

requiredcapacitywhenneeded,andwithoutanyinterruptions,ismostprobablyanindispensable

featureofapplicationsrelyingonthesupportfromtele-operationfromtheSCC.Alsoimportantwillbe

thesecurityoftheSCCs.

Thetransitiontotheautonomousunmannedshippingeracanbehypothesizedtotakeplacegradually

overaperiodofafewdecades.Thefirstapplications,especiallythosebeingupgradesonexisting

vessels,couldbeexpectedtostillcarrysome,althoughreduced,crewonboardforspecifictasksand

availableasinsitubackupincaseofproblemsencounteredatsea.However,singleapplicationswith

evenhigherlevelsofshipautonomycouldalsobeexpectedalreadyinnearfutureinsomelocal

specificservicesespeciallywellsuitedforunmannedoperation.

2. Are‘unmannedships’safe?Thepresentedvisionsoffutureautonomousshipssailingunmannedhaveraisedgenericconcernand

questionsamongsomeprofessionalsandwell-informedlaymenaboutthecredibilityandsafetyof

suchshipsascomparedtoconventionalshipsoperatedbyacrewon-board.Examplesofsafety

concernsexpressedhaveconsidered:

• abilityofautomationtoreliablydetectsmallvesselsandfloatingobjectsonroute;

• abilityofautomationtoavoidcollisionsincaseofencountersofmultipleships;

• abilityofautomationtonavigatesafelyoncoastalfairways;



60

• reductionsonpreventiveandcorrectivemaintenancethatarecurrentlylargelycarriedout

duringvoyages;

• abilitytohandleemergencies,suchasfirefightingorfailurerecoveryandrepairsatsea;

• errorsandmalfunctionsinsoftware;

• disturbances,malfunctionsandvulnerabilitiesindatacommunicationconnections;

• unduetrustonthecapabilityandflawlessnessofICTsystems

Fromasecuritypointofviewconcernshavebeenraisedastothehighervulnerabilityofenvisaged

unmannedshipstohijackingorpiracywiththepurposeofsteelingthecargoorkidnapingthevessel

forransom.SimilartotheconcernsraisedregardingcybersecurityofICTsystemsingeneral,potential

vulnerabilityofunmannedshipstocyber-attacksbydifferentadversaries,allowingthemtoillegally

manipulateorexploittheattackedsystem,hasbeenespeciallyunderlined.Thisstrongconcernreflects

therousedpublicawarenessoncybersecurityandisjustifiede.g.bythebreachesincybersecurity

pointedoutrecentlyonsomeautonomousroadvehiclesandinotherexamplesonsomeotherfieldsof

newtechnology.

Contrarytothefearednegativesafetyandsecurityeffects,claimshavealsobeenmadeforthehigher

safetylevelsofshipswithhigherlevelsofautomationandoperationautonomy.Suchclaimshavebeen

reasonede.g.basedonhighinvolvementofhumanerrorinaccidentsatseainthepast,andthehigh

crewfatalityratewhencomparedtootherindustriesobservedcurrently.Bothoftheseissuescouldbe

hypothesisedtobereducedbyincreasedshipautonomybyreducingthehumaninvolvementindirect

controlofships,andbyreducingthesizeofthecrewon-boardandexposedtohazardsofthehostile

seaenvironment.

Whileaddressedinitiallyinfewstudies,itappearsthattheimpactsofunmannedmerchantshipson

maritimesafetyhavenotyetbeenstudiedcomprehensively.Furthermore,thereisnoexperience

availableonsuchshipsandtheirsafetyineverydayuse.Thereforeitisofhighimportanceinanynew

developmentprojectsthatthesafetyrisksaresystematicallyaddressedfromthebeginning,andthe

knowledgeonsafetyimplicationsaresystematicallybuiltup,withoutforgettingtheapplicable

experiencefromotherapplicationsfromthepast.

3. PreconditionsofsafetyandsecurityIngeneralautonomousandremotelycontrolledshipsfacesimilarsafetythreatstoconventionalships,

i.e.threatsarisingfromtheseaenvironment,othershipsoperatinginclosevicinity,andships’own

operations.Incaseofautonomousorremotelycontrolledships,however,therecognitionofand

responsetothosethreatsistransferred,toacertaindegree,fromtheon-boardcrewtointelligent



61

softwareandsensorsystemsoperatingon-board,ortosupervisorsmonitoringandcontrollingthe

shipsviadatalinksremotelyonshore.Inaddition,theinterconnectedICTsystemsneededforthe

autonomousorremotelycontrolledoperationbringalongnewriskstobeaddressedandmitigatedin

thedesignandcommissioningofthesystems.

Tobesafeinitsoperation,anautonomousorremotelycontrolledshipshallnotproduceasafety

threattoitself,thesurroundingshipsandproperty,orthemarineenvironment.Inaddition,itneedsto

beabletoadjustitsoperationifgettingthreatenede.g.byothershipsorunexpectedchangesinthe

environment.Thisimplies,ingeneral,thatatele-operated,highlyautomated,orevenunmannedship

mustbecapableof:

• generating,oratleastusing,avalidvoyageplanforaforeseenseavoyageandassuringthe

ship’sreadinessforthevoyagebeforedeparture;

• navigatingaccuratelyaccordingtothepredefinedvoyageplan,andavoidingcollisionswith

othertrafficandobstacles–bothfixedandfloating-encounteredduringvoyage;

• maintainingitsseaworthinessandoperabilityoverthevoyageascarriedoutinvaryingsea

states;

• respondingsafelytocriticaleventsandadjustingitsoperationtopotentiallydangerous

changesintheoperatingenvironmentandshipconditions;

• facilitatingemergencyinterventionsforrecoveryandrescueatsea;and

• resistingunauthorizedintrusionsintoshipsystems,eitherphysicalorvirtual,withtheaimof

maliciousactsorillegalexploitation.

Therelativeimportanceofthesedifferentaspectsobviouslydependsontheparticularapplication,i.e.

shiptypeandtheserviceitisproviding,characteristicsoftheoperatingarea,etc.

Ageneralrequirementcommonlystatedforautonomousunmannedshipsisthat,inordertobe

acceptabletocommercialuse,theymustbeapprovedtobe‘atleastassafeastheconventionalvessels

currentlyinuseforsimilarpurpose’.Someclaimshavealsobeenpresentedthatthelevelofriskthat

canbeconsideredacceptableregardingseverecasualtiesshouldbenotablylowerforautonomous

ships.Thiswouldreflecttheassumedlowerpublictoleranceofriskincaseofautonomousshipsdue

toperceivedlowerlevelofcontrolbythepeopleinvolvedonevolutionofsuchsituationscomparedto

conventionalvessels.



62

4. Focalareasofrisk–someselectedexamplesInmarinetechnology,risksareoftenassessedandanalysedbycategorisingthemindifferenttypesof

marineaccidents,like:collision,contact,grounding,fire,explosion,capsizing/listing,flooding,

foundering,hullfailure,lossofcontrol,andinsomeaccidentstatisticsadditionally:unknown.

However,incaseofanew,emergenttechnology,suchanapproachmaynotnecessarilybefruitful

enoughforourpurposes,especially,iftotallynewhazards,risksandriskcontroloptionsneedtobe

identifiedandassessed.Therefore,amoreholisticview,notlimitedbytheconventionalwaysof

thinking,isconsiderednecessarytoobtainamorecomprehensiveknowledgeandunderstanding.

Areviewofrelatedliteratureandpreliminaryassessmentofautonomousandremotelycontrolled

shipoperationpointsoutcertainimpactsofincreasedautonomythatcouldbecomedetrimentalto

safetyofshippingunlessproperlytakenintoconsiderationwhendesigningandimplementingthe

systemsforsuchoperation.Safetyofautonomousshipsdependslargelyonthedesignand

technologicalimplementations.However,inadditiontotheinteractionsofthevariouscomponents

andsub-systemsinthetechnology,humanoperatorsandthehuman-technologyinteractionremain

evenmoreimportantelementsinthisimplementation.Inlinewiththis,theriskscouldbecategorised

intothoserelatedtothetechnologiesneededtoimplementautonomousshipoperations,andthose

relatedtooperatingthistechnologysuccessfullyaspartofthemaritimetransportationsystem.

Selectedexamplesofbotharebrieflydescribedinthefollowingsub-chapters.

4.1 Reliabilityofsafetycriticalequipment

Thereisanincreaseddemandforreliabilityordependability7inshipsapplyinghigherthanusual

levelsofautomation.Asamatteroffactthedevelopmenttowardstele-operatedorautonomousships

demandsdependable,safeandsecuresystemson-board,extendingtotheshipitself,itssystemsand

itsenvironment,includingallservicestheshipsuse.Thus,allsystemstheshipispartof,andall

systemsandsub-systemstheshipuses,areinvolved.

ICTsystems

Shore-basedremotemonitoringandcontrolobviouslyreliesontheexistenceofreliableandsecure

communicationlinksbetweenthecontrolcentreandtheshipsunderitssupervisorycontrolsothat

sufficientspeedandbandwidthfortheneededdatatransferiscontinuouslyavailable.

Inprinciple,afullyautonomousvesselcouldoperatesuccessfullyforalongtimewithouthavingan

operationaldatalinkwiththeremoteShoreControlCentre(SCC).However,ifcontrolbyanSCC

7 Dependability is an umbrella term. It includes several sub-terms: reliability performance, availability performance, maintainability performance, supportability performance, integrity, safety etc. For further details, see IEC TC56 standard.



63

operatorisconsiderednecessaryasanemergencybackup,availabilityofanoperationaldatalink

wouldneedtobeverifiedasaprerequisiteforthevesseltooperate.Inotherwords,startingasea

voyageshouldnotbeallowedunlessadatalinkarrangementhavingsufficientcapacityforemergency

operationsandrequiredreliabilityoverthemissionisknowntoexist.Typically,atleastpartially

redundant,divergentdatalinkstofacilitatetheneedsforcommunicationsinthedifferentoperational

situationswouldberequired.

Similarly,robust,compatibleandproperlyvalidatedICTstructuresandsoftwarearerequiredbothon-

boardthevesselsandattheshorecontrolcentre(SCC)inordertoavoidrisksrelatedtoflawed

operationoftheembeddedsystemintelligence.

Reliabilitymanagement

Conventionalshipsappeartorelystronglyonthecrewon-boardasaninsituresourcefortimely

failurerecoveryatseaandexecutionofpreventivemaintenanceprogramsonlineduringthesea

voyage.Thisallowsusinglesscostlymachineryconfigurationsthatrequirefrequentpreventive

maintenanceactionsandhavelowerreliabilitywithrespecttofailuresrepairableatsea.

Lackofpermanentcrewon-boardwouldessentiallydiminishthecapabilitytoperformpreventiveand

correctivemanualmaintenancetasksonshipequipmentduringseavoyages.Thisimpliesthatsystems

essentialforoperationneedtobedesignedtoberesilienttofailureandextendedmaintenance

intervals.Lackofpermanenton-boardcrewalsocreateshigherdemandsforschedulingof

maintenanceactionsonharbourstays.Thiscallsfortheintroductionofefficientdiagnosticsandnew

predictiveprognosticalgorithmstohelpassessingandcontrollingtheriskoffailuresand

preschedulingofrequiredmaintenanceactionsaspartofoverallshipoperationplanning.Designing

easilymaintainablesystemswouldhelptominimisethetimeandresourcesrequiredandtoassure

thattheactionsarecorrectlyperformed.

Regardingmachinerysystemscontrol,acommontrendseemstobetowardsremotemonitoringand

controlfromshore-basedservicecentresrunoftenbythemanufacturer.Inthiscontext,alsothe

controlofthestatus/healthofotherimportantequipmentthanthemainmachineryneedstobe

maintained.

Basedonexperience,revisionsandrepairsmadeonexistingsoftwareintensivesystemsrepresenta

commonrisktoerrorswithimmediateorlatentimpactsonsystemperformance.Consequently,

revisionsorrepairsonsuchsystemsneedtobethoroughlyplannedandmanagedwithproper

configurationcontrolandcomprehensiveverificationtestingprocedurestosupportrecommissioning



64

ofthesystemsbacktonormaluse.Allchangesandmodificationsshouldbetrackableandthus

systematicallyandtruthfullyregisteredinvessel/companylogbooks.

4.2 Humanfactorsissuesinremoteoperationandmonitoring

Thereareavarietyofpotentialchallengesrelatedtooperationandmonitoringoftheunmannedships

withsafetyimplications.Firstly,theexistingliteraturehaspointedoutthatduetoteleoperationthere

wouldbenobodilyfeelingoftheshiprockingorshipsense.Itisthereforepossiblethatfull

understandingoftheconditionswouldnotbeachievedviacamerasystems.Insmallerships,steering

canbeadjustedinaccordancewiththewaveformationthroughbodilysenseoftheship.

Automationandremoteoperationimpliesthattheshipswillbeequippedandoverviewedwith

multiplesensors.Thedangerhereisthattheoperatorcouldbeexposedtoinformationoverloadand

thereforenolongerabletomakesenseofthesituation.Theproblemwouldbeevengraverifone

personwouldmonitorseveralvesselsassteeringtheoverviewfromonevesseltoanothercouldbea

potentialpointformishaps.WithUASs(unmannedaircraftsystem)severalmishapshaveoccurred

duringchangeoversorhandoffs,thesehavingbeenthedirectorindirectcauseoftheincidents.

Representingseveralsourcesofinformationinoneindicationviaso-calledsensorfusionisapotential

solutiontothisproblem.Thismightbeproblematicaswellasitcanbeimportantfortheoperatorto

understandeachofthesensors.Allofthesensorsmightnotalwaysbeworkingandtheymighteven

provideconflictinginformation.Tofullyunderstandthesituation,theoperatorwouldneedso-called

automationawareness,thatis,comprehensionofthecurrentandpredictedstatusofautomation.Yet,

achievingfullunderstandingonwhatdifferentaspectsofautomationaredoingcanbedifficultifthe

sensordataisfusedtogether.Thisfusionshouldbedoneinamannersuchthatthesystemis

transparentfortheoperatoryetwithoutinducinginformationoverload.

Afurthercomplicationispotentialskillshortage,andskilldegradationatalaterphase.Thefirstissue

hereisrelatedtotheavailabilityofonboardtrainingvacanciesfordeckandengineratingsandcadets,

ife.g.thenumberofcabinsandtrainersonboardgetdiminished.Assumedly,withrelianceon

automationandwithoutmanualdrivingactivity,itisdifficulttomaintainskillsneededinvarying

maritimeactivities.Withrespecttoabnormalsituationsthiscouldbeespeciallydifficult.Maintaining

goodskillscouldbeespeciallydifficultifmonitoringafleetofdifferentkindsofships–theoperator

couldhavetolearnthepracticaldifferencesofeachoftheshipsandcouldeasilyforgetorfailto

recognizerelevantissueswhenswitchingtheoperationfromoneshiptoanother.

Drivingtheunmannedshipsremotelybyteleoperationcouldbechallengingduetolatency.Ittakes

timeforasignaltotravelviasatellitesorothermeans.Thisimpliesthatinteleoperationthereis

alwayslatencypresent.Toomuchlatencycaninhibitactualisingpracticaltasks,i.e.,withtoomuch



65

distancepluslatencytheso-calledcognitivehorizoninteleoperationcouldbeexceeded.Accordingto

research,50msdelaybordersthelimitofdelaydetectionforhumanbrain.Adelayof200msis

consideredtobenoticeableinpractice.

Additionally,boredomhastobeconsidered.Forexample,inapreviousstudy,92%ofUAS(unmanned

aircraftsystem)operatorshavereported“moderate”to“total”boredom.Boredomcouldresultsasa

lossofvigilanceandisthereforeariskfactor.

Asasummaryofmanypotentialhumanfactorschallenges(excludingsecurity)inautomatedshipping

wemaypresentthefollowinglistofissuesthatneedattention:

• Diminishedshipsense

• Informationoverload

• Mishapsduringchangeoversandhandoffs

• Needforautomationawareness

• Skilldegradation

• Latencyandcognitivehorizon

• Boredomandvigilancemaintenance

4.3 Security

Securityreferstounauthorizedintentionalactsofpersonsororganisationsaimedtocauseharmor

damageto,ortoillegally/criminallyexploit,asystemforthepurposesofthemaliciousactor.Piracy,

theftofcargo,smugglingofgoods,humantrafficking,damagingofshiporportfacility,vandalismand

sabotage,hijackingofshiporpersonson-board,useofshipasweaponforterroristactivity,etc.are

commonlylistedexamplesofmarinetransportrelatedsecuritythreats.Aparticulartypeofthreat

beingcredibleforaparticularshipobviouslydependsonhowpotentialactorsperceivethethreattype

andtheshiptomatchtotheirobjectivesandperceivedcapabilitiesforsuccessfullyexecutingthe

plannedmaliciousact.Vulnerabilities(i.e.gapsordefects/weaknesses)identifiedintheprotectionsof

shipsystemscouldbeconsideredasanexampleofpotentialincentivesforattemptingtheactand

selectingtheshipasthetarget.

Theactorsformaliciousactsmaybeexternalto,orcomefrominside,theorganization.Traditionally

executionofmaliciousactshasrequiredphysicalpresenceoftheactorsandintrusionintothetarget

system.ThegrowingusageofnetworkedICTtechnology,however,hasmadeitpossibletotryto

accesssystemsvirtuallythroughnetworkinterfacesandgainunauthorizedremotecapabilityto

manipulateorexploitthesystemoritsparticularelementsinsomeundesiredmanner.



66

Cybersecurity

Thecontinuousincreaseofconnectedon-boardICTsystemstosupportshipoperationsandtheuseof

differenttypesofdatanetworkstomaketheshipsatseaaccessibleforvarioustypesofremote

onshoreserviceshasinitiatedcommonconcernsoncybersecurityofsuchsystems.Inotherwords,

seriousquestionshavebeenraisedwhethertheimplementationsofsuchsystemscanactually

effectivelyresistmaliciousactsonshipsthatmaybecomeattemptedremotelyviatheICT

infrastructures.Thisconcernandprovokedawarenessisreflected,forexample,inIMOsafety

committeeworktopics,specialnumbersonprofessionaljournals,andbytheguidelinesthatLloyd’s

RegisterrecentlyissuedforICTsystems’designandassuranceonships.

Concernsoncybersecurityarefurtherincreasedinthecontextofautonomousandtele-operated

ships,inwhichtheconnectivityofsystemsisfurtherexpandedtoallowtheshipstorunin

autonomousmodeorbeoperatedremotely.Thisimpliesthat,inprinciple,anybodyskilfulandcapable

toattainaccessintotheICTsystemcouldtakecontroloftheshipandchangeitsoperationaccording

tohackers’objectives.Thiscouldmeansimplysomedisruptiveactionsormanoeuvresintroducedfor

annoyanceordemonstration,hijackingoftheshipandcargoforransom,butalsopoweredgroundings

orcollisionscreatedonpurposetocauseseveredestruction.Inadditiontohackingintothesystems,

operationofautonomousshipscouldalsobethreatenedbyintentionaljammingorspoofingofAISor

GPSsignalsorthedatacommunicationsbetweentheshipandtheshorecontrolcentre.

ProtectionagainstcyberthreatswouldcallforeliminationofvulnerabilitiesintheICTinfrastructure

andimplementationofeffectivemeasuresforintrusionprevention,aswellasintrusiondetection,

damagecontrolandsaferecoveryincaseofthepreventionmeasuresfailing.Reflectingthefactthat

potentialattackerswillgetmoreskilfulovertime,andwillhavemoreadvancedtechniquesavailable

tothem,theoversightoncybersecurityneedstobedynamicandproactiveintroducingupdatesinthe

systemsaccordingly.Dataclassification,dataencryption,useridentification,authenticationand

authorisation,dataprotectionagainstunauthoriseduse,dataintegrityprotection,connectivity

protection,andactivityloggingandauditingareexamplesofcommoncybersecuritymethods

foreseentobeneeded.Althoughsomepartsoftheprotectionincybersecuritymaybeautomaticthere

isnodoubtthatasufficientamountofresourcesneedtobeallocatedforthispurpose.Inadditionto

thetechnologyimplementationthelevelofcybersecuritywouldobviouslydependoneducationand

theorganizationalcultureguidingperformanceofthepeopleinvolved.



67

Occupationalsafetyandhealth

Socialsecurityisatopicthathasseveralmeanings.WithintheframeworkofAAWA,wemayinclude

underitscoversomeissuesrelatedtooccupationalsafety,healthandwell-beingoftheseamenon-

boardshipsandashore.Inthisareaautonomousandtele-operatedactionsorshipsmayhelpin

avoidingoratleastreducingoccupationalaccidentson-board.Itmaybepossibletoincludemany

differentsecuritymatters,withinter-activefeedbackmechanismsandimpactsamongthelistsofhigh-

levelinteresttopicsoftheindustrywhennewtechnologyisimplementedon-board.

WithintheframeworkofAAWAsomeissuesalreadyincludede.g.intherelevantconventionsofILO8

mayalsobeseenasindirectproactivemeasuresagainstcyberthreats.Awide,holisticviewmaybring

newaspectsandpointsofviewintodiscussion.Althoughadeeperanalysisofsocialsecurityisleftout

ofthescopeofthisstudy,itmustberecognisedthatmanydifferentconceptsofsocialsecurityofthe

maritimecommunityandsocietyexist,andmanyofthemareofteninterconnected.

4.4 Cargomanagement

Inconventionalships,thefirstofficerandshipmasterareinchargeofacceptingthecargoandits

loadingintothecargospaces.Lackofpermanentcrewon-boardtheautonomousshipswould

emphasisetheroleofportoperatorsinacceptingthecargoandassuringthatitiscorrectlyloadedand

stowedon-boardinaccordancewithshippingregulationsandtheshipspecificcargomanual.

Furthermore,inunmannedshipsandshipsunderremotecontrolpossibleactionstotakeanycargo

relatedmeasuresatseaaremorelimitedthaninconventionalships,ifnoextraequipmentfacilitating

additionalmeasures,e.g.foradditionalcargomonitoring,securingorcontrol,areprovided.

Incaseofautonomousships,assuranceofproperinitialstatusofthecargofortheforeseenseatrip

wouldrelymainlyonthelongshoremen.Thiscouldincreasetheriskofcargorelatedincidents,asitis

believedthatcrewmembersandofficerssailingon-boarddohaveadeeperpersonalinteresttoensure

thatcargoloadingandthesecuringworkaresafelydoneandtheequipmentusedarefitforthe

purposeinallconditions.Anyactionstocurecargorelatedproblemsidentifiedon-board,like:cargo

shift,leaks,problemswithmoisture,fireandfloodingarelimitedtothosethatcanbehandledeither

byautomationortele-operation.

4.5 Managingemergencies

Lackoftrainedcrewmemberson-boardcouldbeexpectedtoincreasetheriskoffailureincopingwith

emergencysituationsthatcanbeencounteredduringvoyages.Decreasedcrewsizemaycreatea

8 International Labour Organization (ILO), see e.g. Maritime Labour Convention, 2006 (MLC, 2006) (with entry into force: 20 Aug 2013), or the earlier Convention C165.



68

higherriskoffailuretohandleemergencieson-boardwhenactionsareneeded.Capabilityofspecific

promptresponseactionsinsituorevacuationofthevessel,ifneeded,hasraisedstrongdoubtsonthe

autonomousoperationconceptbeingapplicableatall.

Howtheshipcanassistinemergencysituationsrelatedtoothershipsisanotherquestionwithseveral

uncertaintiesduetothelimitationinthecurrentlyavailabletechnicalspecification.Emergency

situationsincludeawideareaofpotentialoperationsthatneedtobediscussedinmoredetailwhen

detailedsolutionsareavailable.

However,althoughreducingthecrewsizemightresultindiminishedcapabilitytoassistothershipsin

emergencysituationsbyhands-onhelp,theautomatedshipscouldhavealternativepositive

contributionstoemergencymanagement.Astudysuggeststhat,thankstoincreasedsensordata,the

automatedshipscouldreproduceinformationtoauthoritiesifneeded.Videoandsensordatacouldbe

transmitteddirectlytovesseltrafficmonitoringservices,whichcouldbehelpfulinincreasing

authorities’situationalawarenessinemergencies.

5. ManagingshippingsafetyandsecurityinshortandlongtermManagementofshipsafetyandsecurityintheshorttermmaybeseenmainlyasaprocessthat

requireshavingspecificwell-definedsystematicproceduresappliedintheclassificationandapproval

processes.Suchprocessesaresufficientforassuringsafetyandsecurityofsuchships.However,there

mayalsoappearneedstobeabletoactinunforeseensituations.Whentheoperationofanew,large

andsafety-criticalsystem,suchasamerchantshipisconsideredtobeallowedforevenalimiteduse,a

precautionaryprincipleissuggestedtobefollowed.Whenthesystemcanpassallthechecksandtests

understoodnecessarytoconfirmitssafetyandsecurity,asteptowardsamorecomplicatedsystemor

useinadifferentenvironmentmaybeconsidered.

Intheshorttermthemistakesmadebyautonomoussystemsmaybestillattributedtohumans,asthe

softwareisplannedandproducedbyhumans.However,controlsmustbeinplacetoensurethatno

badorerroneousinformationordistortedideasofthefunctionalitiesorenvironmentalconditionsare

used.

Inthelongertimeperspective,managementofsafetyofautonomousshipscouldbeexpectedthrough

IMOregulationsandconventionsbeingadaptedtobetterencompassalsoautonomousmodesofship

operationandtheassociatedsafetyrisks.Theinitialrisk-basedapproachesforapprovalcouldbe

expectedtodevelopintostandardisedprescriptiveandgoal-basedrequirementstoguidethedesign

andimplementationofautonomousfeaturesonshipsandtheonshorecontrolcentres.Safety



69

managementshouldbedirectedtoacomprehensive,holisticviewextendedtoaspectsandissues

relatedthefulllife-cycleoftheautonomousship.

5.1 Qualificationofnewtechnologiesforuse

Qualificationofnewtechnologies,suchasICTsystemstoenableautonomousortele-operatedships,

forcommercialusecanbeoutlinedasastep-wiseprocess.Smallstepsthatgraduallybuildup

confidenceinthenewtechnologyproposedwithcontinuousimprovementsensuringthatitfulfilsthe

requirementsidentifiedforsafeoperation.Implementationofnewtechnologycanbeseenasa

learningprocessduringitswholelifetime.

Currentlyinternationallyagreedconventions,suchasSOLAS(i.e.SafetyofLifeatSea),specifythe

minimumstandardsfortheconstruction,equipmentandoperationofshipsconsideredtoenablesafe

operationtogetherwithsuchcodesandregulationsasCOLREG9,ISM10andSTCW11etc.These

standardsincludeprescriptiverequirementsonstructuraldesign,specificequipment,sizeand

qualificationsofcrew,etc.,compliancetowhichneedstobeproofedforeachindividualship.Deviation

fromanyprescriptiverequirementrequirestheshipownertodemonstratewithsufficientevidence

thattheproposeddeviationisatleastassafeastheinitialrequirementintheconsideredservice.

Basedonsuchdocumentedevidenceonunaffectedorreducedsafetyrisk,theFlagstatecanthenissue

anexemptionpermitforthedeviantsolutioninaparticularshipandservice.

Thesafetyassuranceprocessforaproposedalternativesolutionneedstostartwithathorough

descriptionoftheshipoperations,bothnormalandabnormal,onwhichtheproposedsolutionis

foreseentobeinvolved,followedbyidentificationofhazardsandothersafetyissuesconsidered

relevanttotheseoperations.Therole,capabilitiesandlimitationsoftheproposedsolutionin

controllingthehazardsandcontributingtotheriskofaccidentsthenneedstobethoroughlyidentified

andassessedtoproduceasuitablebodyofevidencetosupporttheargumentofsafetyequivalenceof

thealternativesolution.The‘standard’solutionscompliantwiththeprescriptiverequirements

providethebaselineforriskcomparison.

Autonomousortele-operatedshipsrepresentamajortechnologicalandoperationalchangewitha

numberofuncertaintiesregardingtheirsafetyinoperation,andnorelevantfielddatacurrently

availabletosupporttheirapprovalforcommercialuse.Wellplanneddemonstratorstudies,carried

outinitiallyatspeciallyplannedsimulatorsettings,andlateron-boardactualseagoingvesselsare

seenasthewayforwardforlearningandbuildinggraduallytheevidenceandconfidenceonsafetyof 9 The International Regulations for Preventing Collisions at Sea 1972 (Colregs) 10 International Safety Management Code 11 Standards of Training, Certification and Watch-keeping for Seafarers



70

suchshipsandtheoperatingconcepts.Obviouslydemonstratorstudieson-boardactualvesselsneed

tostartwithlimitedscopeandendeavour,andhavingacompetentcrewon-boardasabackupand

readyfortakeoverthecontrolincaseofseriousproblems.Theshorecontrolcentre(SCC)work

processesconstitutesanotherareaoftheautonomousshipsysteminwhichdemonstratorstudiesare

seennecessary.

5.2 Managingtherisksduringtechnologytransition

Duetothenatureofshippingindustry,thetransitionfromthecurrentconventionalconceptsin

marinetransportationtoastagedominatedbyautonomous,unmannedshipsisexpectedtotakeplace

slowly,andhasbeenclaimedtorequireatleastacoupleofdecades.Duringthisperiodtherewouldbe

amixtureofvesselswithdifferentlevelsofautonomyoperatingatsea.Intheworstcase,thismaylead

tounexpectedbehaviourofsomesystems,hazards,and,consequentlyrisks.

Oneimportantaspectinthetechnologytransitionisthemanagementofmaintenanceandrepairof

systems,andensuringonlyas-plannedinteractionsbetweene.g.subsequentsoftwaregenerations.

Thewell-performedmanagementwithstandardisedroutinesofup-to-datedocumentationisan

importantpartandfeatureofthesystemicapproach.Theareasofresponsibilityshouldalwaysbe

clearduringallphasesofthetechnologicaltransition.

5.3 Obtainingandmaintainingoperatorskills

ItisclearthatanupdatedtrainingregimeofSTCW(StandardsofTraining,CertificationandWatch-

keepingforSeafarers)willbeneeded,beforeanyfurtherstepsaremadetoallowcrewreductions.The

crewmembersneedtobetrainedinanycasetofulfilallfunctionaltasksandcapabilitiesleftforthe

crewinautonomousships.Itisnotquitecleartoushowthiswillaffectthecrewlists,butatleastin

thebeginningthereisanimportantphase,whentheautomaticoftele-operatedoperationsneedtobe

observedandsupervisedon-board.

SimilartypeofrequirementsasinSTCWmayeventuallyhavetobedevelopedforpersonsoperating

shipsremotely.ItisrecommendedthatpersonsworkingintheShoreControlCentres(SCCs)are

requiredtohaveasufficientamountofexperiencerelatedtosimilarships,i.e.withregardto

dimensions,deadweightandpowerandtheirrelations.Theserequirementsofcompetence,

knowledgeandunderstanding,basedonhands-ontraininginseaserviceandsimulators,shouldbe

clearlyhigherforthesupervisorsintheSCCs.

Goodskillsareneededinsafetycriticalandchallengingsituations.Thereareseveralissuestobe

consideredinobtainingandmaintainingtheoperatorskillsforremoteoperationtasks.Manualskills



71

weakenwhentheyarenotused,thatis,itcouldbeproblematiciftheoperatorusuallyonlymonitors

theshipsandattimestakescontrol.Inremotemonitoringchallengingsituationsseldomhappen,yet

highlevelofcapabilitywouldbeneededinchallengingsituationsinparticular.Workingknowledge

canonlybeachievedthroughrepeateduseofthesystemandiftheworkmainlyinvolvesmonitoring,

thismightnotbepossible.Inthemaritimecontext,thevesselsareusuallyallmoreorlessdifferent.

Theoperatorwouldnotneedtolearntheparticularitiesofalloftheships,butatleastboththeoretical

andpracticalknowledgeandunderstandingofthemaincauseandeffectrelationsandtheirvariations

duetothepeculiaritieswouldbeadvisable.

Overall,welldesignedsimulatortrainingwouldbeneededforpracticingchallengingsafetycritical

situations.Thisisnotunproblematic,since,atleastinprinciple,thesimulatorcannotpresent

unimaginablesurprisingsituations.Creatingchallengingsituationsdemandscreativityand

understandingofmaritimeaccidentsfromthedevelopersoftraining.Theoperatorswouldneedto

havesufficienttrainingdaysatthesimulator,wherethesesurprisingandchallengingsituationswould

takeplace.Debriefingafterthesimulatorsessionsisimportantandneedtobedesignedaswell.In

debriefing,theoperatorshouldbeabletoevaluatehisorherownperformanceandhencewouldlearn

fromsuccessesaswellasfromfailures.Oneoptionistoshowanddiscussvideoclipsofoperators’

actionsduringsafetycriticalsituationsatthesimulator.

5.4 User-centreddesignandvalidationoftheshorecontrolcentreoperation

Inviewofsafety,itisessentialforthedesignofgoodremoteoperationandmonitoringsystemstodo

fieldstudiesonactualmaritimeactivityonregularships.Thestudiedconventionalshipsshould

performthesametasksasthenewunmannedshipswouldactualize.Welldonefieldstudiesallow

understandingsafetycriticalaspectsofworkandmeansformaintainingsafety.Task-analysesbased

onthefieldstudiesallowunderstandingofwhataspectsinactivityshouldbeleftfortheautomation

andalarmsandwhatshouldbemonitoredbyothermeans.Fieldstudiescanalsorevealsurprising

safety-relevantaspectsfromworkers’activities.Knowledgeoftheseallowstakingtheminto

considerationindesign.

Afterdesigningandimplementingthesystem,validationthatittrulyworksaswantedisessential.In

othersafetycriticaldomains,suchasinnuclearpowerplantoperation,thisinvolvescontrastingthe

findingsfromtestingthesystemtosafetystandardsanddemands.



72

6. BuildingriskunderstandingforthefutureAsasteptowardstheeraofautonomousmaritimetraffic,theFinnishAAWAprojectaimstomakeits

shareinbuildinguptheawarenessandunderstandingonsafetyandsecurityrisksrelevanttothe

envisionedautonomousconcepts,andthemeasuresneededtomanagetheseriskseffectively.

Basedoninitialidentificationofhazardsandriskscarriedoutontheconceptofunmannedship,a

numberofriskissueshavebeenpointedoutthatcouldbeproblematicintermsofsafetyorsecurity,

buttowhichwehaveconfidencethateffectivesolutions,riskcontroloptions,canbefoundwith

properlyfocusedandsystematictestingandresearch.Ontheotherhand,thenewconceptsof

operationandthetechnologiesaimedtofacilitateitincorporateissuesthatweknowtoentail

elementsofrisk,buttheseriousnessandthecomplexityoftheriskiscurrentlylargelyunknowntous.

Finally,characteristictoanynoveltechnologyinitsinfancy,wecanassumethatsomeriskissuescould

yetbehiddenorvaguetous.

Theriskknowledgewillbebuiltupgraduallythroughcomprehensiveanalyses,simulatorstudies,

pilotdemonstratorstudiesexecutedonactualseagoingvesselswithsomethoroughlyconsidered

restrictions,andfinallyincommercialuse.Consequently,inthefollowingphasesofAAWAproject,the

mainemphasisregardingsafetyandsecurityissueswillbeplacedon:

• systematicriskidentificationandassessmentfocusingonbothdesignandoperation

interactionsandprocesses;

• assuranceofcyber-security;

• validationofalgorithmsforautonomousnavigation,includingobstacledetectionandcollision

avoidance,andassuranceofeachcorrespondingsoftware,andtheirsafeinteractions,by

analysisandsimulations;

• user-centreddesignandvalidationofshorecontrolcentreoperations;

Asidefromtheseeffortstoimproveknowledgeonrelevantsafetyandsecurityrisksandtheircontrol,

comprehensiveSafetyCasedocumentationwillbecompiledinAAWAfortheplannedfield

demonstratorcase(s)tosupportapprovaloftheexemptionsneededfromthenationalmaritimesafety

authoritiesforcarryingoutthedemonstratorstudies.

7. RecommendationsBasedontheworkcarriedoutsofarintheAAWAInitiativethefollowingrecommendationscanbe

made:

• Remoteandautonomousshipsshallbemadeatleastassafeasexistingvesselswithsufficient

confidence,takingintoaccountrelevantuncertainties,e.g.environmentalconditionsand

disturbances



73

Astheshareofuncertaintyandriskmaygrowatleastinthebeginning,whenmovingtowardshigher

levelsofautomatedroutines,thetargetlevelofsafetymustbesetactuallyhigherthanthecurrentone.

Thereissomepotentialtoreducehumanbasederrors,butatthesametimesomenewtypesof

hazardsandriskmayariseandwillneedtobeaddressed,e.g.intheareaofcybersecurity.

• Progressonlybysmallandcautioussteps

Applicationofthisprecautionaryprinciplemeansacarefulandsystematicapproachinrisk

assessment,design,deploymentandoperations.Thisrequiresincreasingknowledgeand

understandingbyresearchinthe‘unchartedareas’of‘unknown’.Additionally,itisimportanttolisten

toallrelevantstakeholders,improvedisseminationandflowofinformation,inordertoavoid

unacceptableriskandtoconfirmsafety.

• Co-operativeactionsareneededtodevelopinternationalstandardsandguidelinesforthe

maritimeindustry,preferablyinco-operationwithinIMO

Internationalco-operationbetweenthenationaladministrations,classificationsocietiesandother

relevantbodieswithinterestinthefieldtoutilizetheglobalmaritimeknowledgeunderthewide

umbrellaofIMOisrecommendedforthefurtherdevelopment.

Co-operationisnecessarytobeabletocreateacommongroundforacoherent,safeapproachwhen

layingoutthefirstsketchesofprinciplestobefollowedintheprocedurestobeusedguidingand

controllingthetechnicalandoperationalsafetyofautonomousandtele-operatedships.Itwillbeeven

moreimportantifandwhenthefurthernewtechnologicalartefacts,likeautonomousandtele-

operatedshipsstarttointeractwiththeoperationalenvironmentofthemaritimesociety.

Maritimesafetyandsecurityisaverywideanddeepconceptualtopic.Itcanbedividedinmany

sectors,includinge.g.shipsafety,cargosafety,maritimetrafficsafety,environmentalsafety,

occupationalsafetyandsecurity.Therecommendationsabovearepresentedonaverygenerallevelin

anearlyphaseoftheprogressanddevelopmentofautonomousships.Therefore,itmustbe

underlinedthattheaboverecommendationsarenotall-encompassing.ItisbelievedthatAAWA

Initiativeactsinchangingshipping,butnotinisolationfromthewholemaritimesectorandsociety.

ThelatterhavemanyfeaturesandactorsgivingfeedbacktoandshapingAAWA,too,tosomeasyet

unknownamount.So,thedevelopmentofthenewtechnologywilloccurinmanyinteractions,known

tobetypicalforthesocialconstructionofallsignificanttechnologies.Allthisneedstobetakeninto

account,whensafeandsecuretechnology,dependingbothondesignandoperationisdeveloped.



74

BusinessJouniSaarni,DevelopmentManager,TurkuSchoolofEconomics,Universityof

Turku

SiniNordberg-Davies,DoctoralCandidate,TurkuSchoolofEconomics,

UniversityofTurku

HannuMakkonen,Docent,SeniorResearchFellow,TurkuSchoolofEconomics,

UniversityofTurku



75

FromInnovationstoMarkets–RedefiningShipping

1.Redefiningshipping–atransitiontoautonomousshipping

Autonomoustechnologieswhichenhanceself-guidingcapabilitiesoftechnicalsystemshavereceiveda

considerableamountofattentioninvariousdifferentindustries.Themarinesectorisnowfollowing

suit.Whiletheconceptofacompletelyautonomousshipmaybecontroversial,itisnevertheless

undeniablethattheshippingsectorisfacingconsiderablechangesasdigitalisationgraduallysweeps

overthetechnologicallandscape.Thetechnologicalchangeisconnectedtoasocialone,astheAAWA

slogan“redefiningshipping”suggests;autonomousshippingisnotmerelyabouttechnologybutalso

abouttherespectivesocialchange.Figure1synthesisesthekeyelementsintoaperspectiveon

autonomousshippingintheAAWAproject.

Figure1.AAWA–Redefiningshipping

Thecentralpanelofthefiguredescribesthelevelsofinnovationforautonomousshipping:single

innovations,combinatoryinnovations,systemicinnovation.Thesingleinnovationsrepresentthe

productinnovationscreatedfromkeytechnologies,forexample,cameras,radars,andothertypes

TECHNOLOGICAL CHANGE

SOCIAL CHANGE

Single innovations

Combinatory innovations

Systemic innovation

TECHNOLOGICAL OPPORTUNITIES

NEW NEEDS



76

ofsensorsthatcompriseacombinatoryinnovationofasituationalawarenesssystem(seeArthur

2009).OthersuchtechnicalcombinationstobedevelopedinAAWAincludee.g.navigationalsystems

orcommunicationslinks.Thesekindsofbuildingblockscomprisethekeytechnologyareasfor

systemicinnovation,i.e.theconceptofanautonomousship.TheoutskirtsofFigure1describe

technologicalandsocialchangesthatcompriseamutuallyfeedingloop:technologicaldevelopment

producestechnologicalopportunities,whereassocialchangealtersthesociallandscapegenerating

newneeds.Themutuallyreinforcinginterplaybetweenthetechnologicalandsocialchangeproducesa

socio-technicaltransitionthatdescribesastateinwhichthenewtechnologicalopportunitiesand

thoserelatedneedsarematerialisedintopracticeandputintouse.Forexample,containerisationas

wellasthedevelopmentfromsailstosteamenginesandfurthertodieselenginesdescribessuch

socio-technicaltransitionsinthemaritimesector(seeGeels2002).*

Figure1connectsthelevelsofinnovationwiththetechnologicalandsocialchangeintwoways:

1. Theongoingtechnologicaldevelopment(digitalisation,theinternetof

things,autonomousdrivingetc.)andrespectivesocialchangeinwhich

thesetechnologiesbecomesociallyacceptedanddesiredfeedthe

innovationactivityinthefieldofautonomousshipping(thespinofthe

outercircleacceleratesthespinoftheinnovationgears).

2. Theinnovationactivitiesforautonomousshippingacceleratethe

“AllthispositivepublicityandenthusiasmthatyoucanseearoundGooglecaretc.ishelpingoureffortsinthemaritimesector.”

Lessons learned from past socio-technical transit ions * The past socio-technical transitions in the maritime sector have some joint similarities. Firstly, the transition usually lasts a long period of time, often decades. Ships have large long-term investments involved and they are constantly a part of transportation functions of world trade. Secondly, the transition typically begins from small special markets. Some combinations of tasks, cargo and routes fit well with the novel technologies that are emerging in the beginning of the transition. Thirdly, it is typical in a transition phase that the old existing regime and new entrants co-exist at the same time and compete. Later, market selection occurs favouring a so-called dominant design from different technological alternatives (see Murmann & Frenken 2006). When the selected technology diffuses within markets and users it also causes social impacts. More concrete targets are the needed infrastructure and service networks for e.g. maintenance. Also regulations and policies might face changes as well as industry structures. More vaguely traceable are institutional and cultural dimensions on routines, practices and mind-sets. (Geels 2005.) Taking the history of containerisation as an example, the social factors were of the utmost importance. Containers and container ships were technologically rather straightforward. More importantly, they were process and organizational innovations. The idea of using containers to streamline the loading and unloading was already variously experimented during the 1950s. However, true utilisation of them needed a change in thinking from operating ships to transportation chains. As a consequence, shipping and port operations became more capital-intensive. This provoked social resistance, and contractual negotiations on container sizes and standards took many years. It took over a decade until the late 1960s for the first purpose-built containerships to be completed. After this, containers’ diffusion in shipping companies and ports with related investments and practice-building lasted for several decades. (Levinson 2006).



77

technologicalandsocialchangeingeneral(thespinofthegearsacceleratesthespinoftheouter

circle),andthusreinforcethecross-industrialsocio-technicaltransitiontowardsautonomous

technologiesandtheirapplicationinthesociety.

Thebusinessmodelsofthekeyactorsintheshippingsectormediatetheconnectionbetweenthe

innovationactivitiesandthetechnologicalandsocialchange.Theinnovationactivityisdependenton

theextenttowhichthekeyactorsperceivebusinessopportunitiesregardingautonomousshipping.

Thisisagainrelatedtotheissueofhowotherkeyactorswillbemobilisedtothetopic,andwhatkind

ofrelationshipsandnetworksaretoemergetoadvancethetechnologicalandsocialchange.Thus,

autonomousshippingislargelyasocialissueinwhichtheprevailingnormsandroutinesofthe

shippingbusinessthatpromotestagnationaretobeovercome.Furthermore,onasocietalleveldigital

solutionsneedtobeseenasmeasuresforimprovingthequalityoflifeinsteadofthreateningit.Based

onthefindingsofthefirstphaseoftheAAWAproject,autonomousshippingisnotaquestionof

whetherornot,butratheraquestionofwhen.

2.Autonomousshipping–anissueofbusinessrelationshipsandnetworks

Duetoitssystemicnature,theemergenceofautonomousshippingisfirstandforemostanissueof

managingtherelevantrelationshipsandnetworks(Håkansson&

Snehota,1995;Håkansson,Ford,Gadde,Snehota,&Waluszewski,

2009),andtheecosystemsbasedonthese.Intermsoftechnology,

theshifttowardsaneraofautonomousshippingrequires

convergenceoftherelevanttechnologies.Similarly,intermsof

themarketsideforautonomousshippingthisshiftrequiresthat

autonomousshippingisperceivedtodelivertheexpected

benefits:onthemicrolevel;seafarersexperiencethattheir

workingconditionsareimproved,onthemesolevel;marineindustryplayersseecost,efficiencyand

safetygains,andonthemacrolevel;thesocietybenefitsfromtheredivisionofworkandlowered

emissions.Thecombinatorialdevelopmentbetweentechnologiesandmarketswithinthemarine

industryoccursinconjunctionwithsimilardevelopmentinotherrelevantindustries(e.g.automotive

andaviation),thataltogethercompriseacross-sectoralautonomoustechnologiesecosystem(as

describedinFigure2).Technologicalinnovationsandthedawnofthenewconceptofautonomous

shippingmotivatestheactorstodevelopnewbusinessmodelsandfurther,newbusinessmodelswith

intentionalactivitiestogeneratecommerciallyviableapplicationsfeedthetechnologiestodevelopas

describedinFigure2.

“Increasingly,allthesesystemsarediscussingwitheachothersomuchthatthediscussionisnolongerbetweentwosystems,butallsystems.Andthisrequiresacertainlevelofnetworking.Butonwhatlevel?It’snotasimplisticsituationinfact.”



78

Figure2: Thenetworksandrelationshipsforautonomousshippingbusiness

Figure2describestheautonomousshippingbusinesstoemergeasaresultofmatchingevolvingneeds

andevolvingtechnologieswithinstrategicrelationships,localnetworksfortechnologyplatforms,and

globalnetworksfornewmarkets.Currently,thedevelopmentofsuitabletechnologiesforautonomous

shippingtakesplacelargelyinstrategicrelationships;businessactorsdevelopsolutionstoservetheir

currentbusinessintheirkeyrelationships.Thiscanbeexemplifiedbyasatellitecommunicationsfirm

workingwithcurrentsuppliersandcustomerstoachievesafercommunicationlinks,oravideo

technologyfirmworkingwithalgorithmspecialiststodevelopthecomputationalpowerofvideo

cameras.



79

Secondly,industry-levellocalnetworksexistinwhichactorshavecometogethertobuildtheconcept

ofautonomousshippingintentionally,asexemplifiedbydevelopmentnetworkssuchasAAWA.These

localnetworkstrulybegintoquestiontheprevailing,traditionallogicofshippingandaimatanew

dominantlogicbyprovidingatechnologyplatform,onwhichfuturedevelopmentcantakeplace

(compareFrenken2000).Whiletheactorsinalocalnetworkmayhaveexpectationsforshort-term

returnsbasedonsingleinnovations,autonomousshippingtrulybecomesadrivingforcefor

innovationinlocalnetworks.Thisdevelopmenttowardsautonomoussolutionshasprogressedfurther

inotherrelevantindustries.Forexample,theautomotiveindustryhasforalongtimedeveloped

technologies(e.g.cameras,radar,ultrasonicsensors)thatformthebasisforintentionalactivitiesto

developandlaunchautonomousdrivingplatforms,suchasthoseofGoogleandTesla.Thesecurrently

representlocalnetworksbutaretransformingtowardsglobalnetworksnotonlycomprising

development,butalsoproductionanduseofcommerciallyviableapplications.

Itisonlywhenotheractors,e.g.customersandotherstakeholdersinthelogisticschain,understand

theapplicabilityofautonomousshippingfortheirneeds,thedevelopmentofautonomousshipping

becomesamatterofincreasedvalueinsteadofafeatofengineering.Atthispoint,thelocalnetworks,

i.e.hotspotsofautonomousshipping,graduallygainmomentumandscaletogenerateglobal

networksengagedintothedevelopmentandoperationofautonomousshipping.Itisintheseglobal

networksthatautonomousshippingeventuallyevolvesintothenewdominantform,i.e.redefine

shipping.

3.Autonomousshipping–arenewedsetofrolesbetweenthekeyactors

Autonomousshippingwillleadtoanewkindofrole-setanddivisionofworkbetweentheactorsinthe

shippingsector.Someoftheserolesareplayedbythetraditionalplayersandsomebynewentrants.

Forautonomousshippingtherewillbenewfunctionsandrespectiveactorswhospecialisein

technologiesenablingthesefunctions,e.g.aremotecontrolcentreoperatorandanautonomous

systemsintegrator.Eachactormustconsidertheirpositioninthemarketrelativetotheotherplayers,

meaningthatactorsshapetheirbusinessmodelsaccordingly.Holdingakeypositioninthetechnology

platformforthenewdominantlogicofautonomousshippingiscrucialforcompetitiveadvantage

(Makkonen,Vuori,Puranen,2016).Asglobalnetworksemerge,moreandmoreactorsengagebothin

thetechnologicalframeworkaswellasusageinautonomousshipping.Thiswillaltertheprevailing

structuresandprocessesoftheshippingindustryinitsentirety.Inotherwords,theautonomousshift

willnotonlystreamlinetechnology-relatedoperationsbutmorewidelyfacilitateacriticalevaluation



80

andreorganisationofthewaytheshippingbusinessoperates.

Figure3describesthepossibleentranceofnewactorsaswellasthechangeoftherolesofthecurrent

actorsintheshippingbusiness.Theincreasingintelligencethatcomesalongwithautonomous

shippingislikelyto1)bringinnewactorstothefieldofautonomousshippingbusinessaswellas2)

bringinadifferentphilosophyintermsofmaintenanceandservicefunctions.Intermsoftheformer,

newtechnologiesdevelopandtechnologicalpotentialmaterialisesinapplicationsoriginallydeveloped

inotherareas,whichcanservetheemergenceofautonomousshipping,asdemonstratedbye.g.the

developmentofdroneaircraftandsemiconductors.Intermsofthelatter,evenafterthedesignand

productionofanautonomousship,newcapabilitiescanbeaddedduetothedigitalnatureofkey

systems.Inthissense,theonceproducedsolutionsandsystemsareneverreallycomplete(Yooetal.

2012),andthusthenetworkofactorsandfunctionsarelikelytobeincontinuousevolution.

4.Transitiondriverstoautonomousshipping

Theanticipatedbenefitsandchallengesofautonomousshippingtobusinessescanbebroadlyviewed

fromtheperspectivesofshippingcompanies,existingmaritimesystemandservicesuppliers,and

possiblenewsuppliersenteringthemarket.

Fromtheperspectiveofshippingcompanies,indiscussionswith

theindustrybothdirectcost-reducingbenefitsandotherindirect

benefitshavebeenpointedout.Thedirectbenefitsareoftenlistedina

“Theindustryneedsnowtostartsearchingforassignmentswhereanautonomousshippaysoffexceptionallywell.”

Figure3Therenewedsetofrolesfortheautonomousshippingbusiness



81

vessellevelasmoreefficientuseofspaceinshipdesign,moreefficientuseofcrewandtheirskills,and

moreefficientuseoffuel.Shippingcompaniesarealsolikelytobenefitandseenewrevenuesfrom

tailoringtransportationchainswithautonomousapplicationsaswellasfromincreasedcargospaceon

theship.Indirectbenefitsareactualisedmoreinacompanyandnetworklevelsoftheshippingsector.

Autonomousshippingallowsimprovedoptimisationofoperationsandprocesses.Forexample,

optimisingoperationsbasedonreal-timedataenableseconomiesofscaleatfleet(orcompany)level,

andreducesthelikelihoodofhumanerrorscontributingbothtosafetyandservicequality.TheAAWA

teamseesthattheseindirectbenefitsarethekeyforgaininglong-termcompetitiveadvantagesfrom

autonomousshipping.

Radicallyrethinkingoperationswithremoteandautonomoussystemsisdeemedtobehinderedby

thecurrentregulatoryenvironment,causinguncertaintyintermsofbeingamongthefirsttoengagein

autonomousshipping.Modifyingregulationforremoteandautonomousshippingisabroadtask

becauseitisacombinationofbothnationalandinternationalrules.Nevertheless,rulesreflectthe

socialopinions,andifautonomousshippingisseentoofferbenefits,itwillgraduallychallengethe

prevailingrules.

Fromtheperspectiveofmaritimesystemandservicesuppliers ,autonomousshippingcan

bringmorepossibilitiesindesigningforimprovedshipefficiency.Supplierscanalsobenefitfrom

significantnewbusinessopportunitiesparticularlyregardingdata-relatedservices.Byengaginginthe

developmentofautonomousapplications,supplierswillgainnewcapabilities,whichcanbeleveraged

bothinfindingnewbusinessopportunitiesaswellas

improvingandbuildingupontheirexistingofferingsinthe

shortterm.Thislearningisenhancedbytheincreasingcross-

sectoralcooperationthatistakingplacearoundautonomous

shipping.Knowledgeandskillsaswellastechnologiestravel

acrossindustryborders,whichsupportstheemergenceofautonomoussystemsnotonlyinthe

maritimesectorbutalsoine.g.theautomotiveandaviationsectors.Theinnovationeffortsofmaritime

systemandservicesuppliersaregraduallysupportedbytherelatedsocietalacceptanceof

autonomoussystemsoverall.Furthermore,regulatorybodiesindifferingflagstatesarealsoshowing

increasinginterestinbackingthecreationofcooperativenetworksthatpursuethedevelopmentof

autonomousshipping.

However,thecurrentregulatoryenvironmentcanbeanobstaclefordevelopingnewbusiness

solutionsinparticulararoundremotecontrol.Whileremotecontrolisregardedasanareawithhigh

potentialforrethinkingoperationswithoutmuchcompetition,thereasonforlowlevelsofcompetition

“Personally,thefirm,thecommunity,theindustry,thesociety,everyonebenefitsfromexchangingexperiencesandthoughtsonsomelevel.”



82

comesdowntoregulationsgoverningcertainpracticesthatcouldotherwisebehandledremotely.

Questionsarealsoraisedwhethertheconservativenaturedmaritimeindustrywouldbereadyto

adoptautonomoustechnologieswiththesamespeedastheybecomeavailable.Also,autonomous

shippingisrecognisedtohaveamajorimpactonthebusinessmodelsofsupplierswhosecurrent

modelsarebuiltaroundtheshippingoperationsoftoday.Reluctancetochangetheprevailingbusiness

modelsmayhindertheemergenceofautonomousshipping.Lastly,uncertaintysurroundingliability

issuesneedstoberesolvedbeforecommercialisationofautonomoussolutionsispossible.Thus,

leapingintothebusinessofautonomousshippingwouldrequirecertaintyoftheinsurers’willingness

tocooperate.

Autonomousshippingcanpavethewayfornewsupplierstoentertheindustry,inparticularfrom

sectorswherethenecessaryhardwareandsoftwaretechnologies(e.g.differenttypesofsensors,data

analyticsetc.)arealreadyinuse.Forthesetypesofentrants,autonomousshippinghasthepotentialto

uncovernewglobalmarketsinshipping.Increasedshipintelligenceopensupnewservice

opportunitiesinparticularforsuppliersspecialisingindataandsoftware.Atthedawnofadigitalera,

theindustryislikelytoseethearrivalofastartupsceneenrichingtheindustry’ssoftwarecapabilities.

Furthermore,duetolowerdemandsforreactiontime,ashipcanbealessdemandingplatformforthe

performanceofmanysystems,incomparisontoe.g.carsandairplanes,makingshippingamore

lucrativesectorfordevelopmentefforts.

Despitetheopportunitiesthatautonomousshippingcanprovidenewsuppliers,shippingis

neverthelessatoughbusinessenvironmenttoenterinto.Forexample,certainequipmentdeveloped

foruseonlandmayfacedurabilityissuesatseaifnotadaptedto

theseaconditions.Itmaybedifficulttoentertheindustryasa

newplayerduetotheinvestmentrequiredtogetallthe

equipmentapprovedtobeabletobringthemonaship,thus

offeringportfoliosforthemaritimesectorneedtobecarefully

planned.Also,buildingbusinessrelationshipsintheconservativeindustrymaybealengthy

endeavour.Furthermore,themarineindustrycouldbeconsideredtobecompetingforknowledgeable

suppliersagainstotherindustriesthataremoreadvancedintheirstepstowardsautonomy,andthus

commercialisationofsolutions.Themaritimeindustryisatadisadvantageintermsofunitvolumes

whencomparedtotheautomotivesectorinparticular.

5.Transitionroadmap

Basedonfindingsmadeonacademicliteratureoninnovations,marketsandsociotechnicaltransitions

supportedbypreliminaryinterviews,asketchinFigure4wasmadeonhowtheautonomous

“It’saveryconservativeindustry,sobeforeyouhaveestablishedanameandtrustonthecustomerside,ittakesalongtime.”



83

transitioninthemaritimesectormighttakeplace.Thefigureisnotmeanttobepredictiveordering.

Rather,itisatoolforunderstandingthedifferenttriggersneededforthetransitiontoproceed

forward.ThebasicconceptinFigure4isthatthesocialacceptanceforautonomousshippingaffects

whatinnovationsareadoptedandtakenintouse.Complexsystemicchangesuchasautonomous

shippingcannotbeimmediatelyadoptedasawholebutitisratherapathandachainofinterrelated

events.

Atthefirststagethesocietyneedstorecognisethattheconceptofautonomousshippingisatleastin

someformpossibleandimaginable.Societyismeanthereinawidesensecoveringalsotherelevant

industryplayersbutalsoauthoritiesandthegeneralpublic.Toseparatearealisableideafromscience

fictiontherecognitionphaseinvolvesconditionsthatsignalthedifferentactorsthattherearerealistic

underlyingpossibilities.Knowledgefromexistingtechnicalperformancebothinmaritimeandother

sectorsshapestheboundariesofwhatisthoughtaspossible.Existingtechnologieslikedynamic

propulsionsystemsorhigh-speedsatellitecommunicationsgiveatestedgroundonwheretothinkup

morefunctionalities.Professionalsarenaturallymorefamiliarwiththetechnicaldetailsbutashared

understandingofprevailingtechnologicalcapabilitiesneverthelessexistsinthesociety.Recentlythe

automotivesectorhasbeenfeedingnumerousexamplesfromautonomousdevelopment.Besides

technicalaspectslikesensorsitalsobringsupmorecomplexthemesintodiscussion.Therearemore

andmorestoriesonregulationorethicsofrobotics.Herethemediahasanimportantroleinspreading

thewordaroundandchallengingexistingthoughts.Theawarenessand“buzz”isculminatedinthe

formofR&DprojectslikeAAWA.Theirtaskistoexplorethenearbyhangingopportunitiesandtake

stepsforwardinthesectorinquestion.AAWAitselfisoneelementcontributingtorecognitionof

Figure4.Transitionroadmaptoautonomousshipping



84

autonomousshipping.Wideinterestinautonomousshippingisalreadyshowingthatrecognitionhas

beenprogressingfastanditisnowtimetomoveforwardinthefigure.

Aftertherecognitionsomevisionaryindividuals,theearlyadopterswillseetheconceptalsonotonly

aspossiblebutalsoasdesirable.Astheoriesforsystemicinnovationssuggest,theinnovationsin

shippingfollowanincrementalapproach.Ashipwon’tbemadedirectlyasfullyautonomousbut

ratherinsmallersteps.Firstconcreteofferingsmightforexamplebenewkindsofdecisionsupport

systems.Theywillbebuiltoverexistingofferingsforshipnavigationandmanoeuvring.Aprecondition

forthatisasituationalawarenesssystemthatmustbedevelopedwithsuitablesensorcombinations.A

decisionsupportsystemwillofferabasiclevelofautonomouscapabilitiesforenhancedobservation

andself-guidanceindifferentoperationalscenarios.Evenifitisnotnecessarilyinconflictwiththe

regulation,basedonthesystem’scapabilitiesandperformance,changestonationalregulationmustbe

evaluated,sothatthedecisionsupportsystemcouldbebroughtintomarkets.Animportant

characteristicinautonomousshippingisthatpartialinnovationscreatedhaveagoodchancetospin-

offtomarketsalreadyalongtheway.

Atathirdstagemoredetailedplanningoccursandinsteadofasingleshipmoreattentionisgivento

managementoftrafficinanautonomousera.Inearlydevelopmentsitistypicalthatmanydifferent

technologicalalternativesemergeandcompetetogether.Eventhoughthisacceleratestechnical

experimenting,atsomepointaneedrisestostandardiseprocedures.Userstendtowantaunified

conceptfortechnologicalartefacts.Thisalsomeansthatpeopleneedtocognitivelyhaveashared

understandingofwhatismeantbytheautonomousshipconcept.Itmustbefairlysimilarastohow

today’sconventionalshipisunderstoodtoholdacertainhomogeneitybetweendifferentships.

Discussionswillstartregardingwhatwillbethestandardsofautonomousshippingindifferent

dimensions.Asthetechnologiesbecomeintowiderusetheremustbesuitableinfrastructureinplace

tosupportit.Shipconnectivitynetworkbasedonsatelliteandshore-basedcommunicationsaretobe

graduallyenlargedtosupporttheautonomoustraffic.Finallystandards,infrastructureandtherising

needfordoingbusinesswiththenewtechnologywillaffectinternationalregulation.Asshippingis

global,thestandardsandrulesmustbeinternationallyagreedupon.Naturallyitisalongprocess,

howeveritisconstantlyreflectedwithnovelapplicationsbroughttobecontestedbythemarketand

bysocialacceptance.

Whentheessentialstandards,rulesandelementsofinfrastructurehavebeensettledthedevelopment

becomesaquestionofexpanding.Therecanbealreadysomespecialmarketnichesthatshowthenew

technologiestoproduceverygoodoutcomes.Thesenichesstarttoexpandintheirownregionsand



85

markets.Graduallysomeregionalseaareaswillbecomeharnessedtohaveafull-scaleinfrastructure

forautonomousoperations.

Inthefinalstageofthefiguretheregionalandniche-specificprocesseswillstarttoaccumulateintoa

globalscale.Autonomousapplicationsdiffuseinsidetheinnovation,reachingapointwhereithasa

significantimpactonglobalshippingandworldtrade(compareGreve2009).Thiswillaltertheroles

andstructurebasedonconventionalshipping.Dataandserviceswillbringmorevaluetocustomers.

Transportationchainswillallowmoreoptimisationandwillbemoretailoredforspecificneedsof

differentindustriesandcustomers.Managementofautonomousfleetsmightconsolidateandglobal

remotecontrolcentresforthesekindsofshipsarebuilt.Ashiftfromproductinnovationstoprocess

innovationsoccurstostartproducingautonomousofferingsmoreefficiently.Onasociallevel,regime

ischanged.Regulation,routinesandpracticesgraduallyincorporatestheautonomousshippingas

takenforgranted.Institutionsofshippingincludethefeaturesandinfrastructurederivedfrom

autonomoustechnologies.

6.Conclusion

Thetransitiontoaneraofautonomousshippingisamorecomplexmatterthanameretechnological

invention.Therealisationofanautonomousshiprequiresaplethoraoftechnologiestobeintegrated

systemically,whichmeansthatcooperationisrequiredbetweenvariousactorswhocanmasterthe

differenttechnologicalareas.However,engaginginsuchinnovativeeffortsmustrealiseabusiness

casefortheactorsinvolved–bothintheshortandlongterm.Thus,benefitsmustberealisedalready

beforeautonomousshippingcanbecomethenorm.Oftenthismeansneworimprovedofferingsinthe

shortterm,whichcanberealisedwiththenewskillsthatactorslearnduringthecooperative

innovationprocess.Assuch,thebusinessaroundautonomousshippingisbuiltiterativelyasaresultof

thecontinuousdevelopmentofsub-componentsthattogethercomprisetheautonomousshipandthe

technologiesneededforitsoperation(e.g.theremotecontrolcentreandcommunications

technologies).Yettheviabilityofthenewbusinessarearequiresactorswhoseinputmakesthe

operationspossible.Theseincludee.g.regulatorybodies,insurers,classificationsocieties,ship

managers,shipowners,shipyards,etc.Moreover,viableshippingbusinessalsorequirescertainnorms

tobebroken,e.g.themarineindustryneedstoovercomeitsconservativenatureifitistobenefitfrom

newsolutions,andthesocietyneedstoacceptdigitalsolutionsasimprovingthequalityoflifeinstead

ofthreateningit.Inotherwords,tofullyrealisethepotentialofautonomousshipping,thedeveloped

technologiesmustbedeemedvaluablebythewidermarineindustryaswellasthesocietyasawhole.



86

Table1presentsthewayforwardtoautonomousshipping.Insummary,autonomousshippingis

possiblefromatechnologicalperspective,astherespectiveindustrialnetworksandsuppliersare

increasinglybecomingorganisedtomaketheconceptareality,andthevalueofautonomousshipping

hasbeenrecognisedfordifferentactorsinthemaritimeindustry.Evenissuesregardingtheregulatory

environment,theoftenpointedoutbarriertoautonomousshipping,appeartobesolvableifthereis

politicalwill.Astheelementsforautonomousshippingarecomingtogether,attentionshouldbe

turnedtohowpeopleperceiveautonomousshippinginthesocietyandtheindustry.Autonomous

shippingmustbecomeculturallyrecognised,anditneedstobecomeanappropriatenorminthe

industry.Suchchangesinmind-setsdonothappenovernight,butthereisindicationthatchangeis

takingplaceasattentionandwiderpublicdiscussionaroundautonomousshippingisincreasinglyon

therise.

Toprovidethemaritimeindustrywithfurtherunderstandingofthecomplexphenomenonof

autonomousshipping,researchersatTurkuSchoolofEconomicscontinuetheirworkin2016-2017in

closecooperationwithindustryrepresentatives.Researchwillincludeexploringthewidermarine

stakeholderperspectivesviaastakeholdersurvey,andinvestigatingthenewbusinessmodelsofkey

actorsintheautonomousshippingecosystemthroughworkshops,enrichedbyinsightsfromother

relevantindustriessuchasautomotiveandaviation.

Table1:Thewayforwardtoautonomousshipping



87

References

Arthur,W.B.(2009)TheNatureofTechnology:WhatItIsandHowItEvolves,NewYork:FreePress.

Frenken,K.(2000)Acomplexityapproachtoinnovationnetworks.Thecaseoftheaircraftindustry.

ResearchPolicy,29(2),257–272.

Geels,F.W.(2002)Technologicaltransitionsasevolutionaryreconfigurationprocesses:amulti-level

perspectiveandacase-study.ResearchPolicy,31(8-9),1257–1274.

F.W.Geels.(2005)TechnologicalTransitionsandSystemInnovations.Cheltenham,UK:EdwardElgar.

Greve,H.R.(2009).Biggerandsafer:thediffusionofcompetitiveadvantage.StrategicManagement

Journal,30(1):1–23.

Håkansson,H.&Snehota,I.(Eds.).(1995).Developingrelationshipsinbusinessnetworks.London:

Routledge.

Håkansson,H.,Ford,D.,Gadde,L.-G.,Snehota,I.,&Waluszewski,A.(2009).BusinessinNetworks.

Chichester:JohnWiley&Sons.

Levinson,M.,2006.TheBox:HowtheShippingContainerMadetheWorldSmallerandtheWorld

EconomyBigger.PrincetonUniversityPress,Princeton,NJ

Makkonen,Hannu–Vuori,Mervi–Puranen,Miia(2015)Buyerattractivenessasacatalystforbuyer-

supplierrelationshipdevelopment.IndustrialMarketingManagement,InPress,onlinepublication.

Murmann,J.P.&Frenken,K.(2006)Towardasystematicframeworkforresearchondominant

designs,technologicalinnovations,andindustrialchange.ResearchPolicy,35(7),925–952.

Yoo,Y.–BolandJr.,R.J.–Lyytinen,K.–Majchrzak,A.(2012)OrganizingforInnovationintheDigitized

World.OrganizationScience,Vol23(5),1398-1408.

© Rolls-Royce plc 2016The information in this document is the property of Rolls-Royce plc and may not be copied, communicated to a third party, or used for any purpose other than that for which it is supplied, without the express written consent of Rolls-Royce plc.While the information is given in good faith, based upon the latest information available to Rolls-Royce plc, no warranty or representation is given concerning such information, which must not be taken as establishing any contractual or other commitment binding upon Rolls-Royce plc or any of its subsidiary or associated companies.

www.rolls-royce.com/marine

Company InputRolls-Royce System Integration and Automation ControlDeltamarin Ship DesignInmarsat Satellite Communications TechnologyDNV GL Classification and Regulatory GuidelinesNAPA Software House providing solutions for Ship Design and Operation

Advanced Autonomous Waterborne Applications (AAWA) partners

Universities InputAalto / VTT (Technical Research Centre of Finland) Safety and SecurityTampere University of Technology / University of Turku Technology ResearchUniversity of Turku Business AspectsÅbo Akademi / University of Turku Legal Aspects

AAWA project coordinationMarkus Laurinen [email protected]

AAWA paper master v12

Documents

Transcript of AAWA paper master v12