AUTONOMOUS VEHICLES THE EMERGING LANDSCAPE An Initial ... Agenda - Autonom… · motor-park...

1

AUTONOMOUS VEHICLESTHE EMERGING LANDSCAPE

An Initial Perspective

Abbreviation DefinitionACC Adaptive Cruise Control - Adjusts vehicle speed to maintain safe distance from vehicle aheadADAS Advanced Driver Assistance System - Safety technologies such as lane departure warningAEB AutonomousEmergencyBraking–DetectstrafficsituationsandensuresoptimalbrakingAUV AutonomousUnderwaterVehicle–SubmarineorunderwaterrobotnotrequiringoperatorinputAV AutonomousVehicle-vehiclecapableofsensingandnavigatingwithouthumaninputCAAC Cooperative Adaptive Cruise Control – ACC with information sharing with other vehicles and infrastructureCAV ConnectedandAutonomousVehicles–GroupingofbothwirelesslyconnectedandautonomousvehiclesDARPA USDefenseAdvancedResearchProjectsAgency-ResponsibleforthedevelopmentofemergingtechnologiesEV Electric Vehicle – Vehicle that used one or more electric motors for propulsionGVA Gross Value Added - The value of goods / services produced in an area or industry of an economyHGV HeavyGoodsVehicle–EUtermforanytruckwithagrosscombinationmassover3,500kg(sameasUSLGV)HMI HumanMachineInterface–Userinterfacebetweenavehicleandthedriver/passengerIATA International Air Transport Association - Trade association of the world’s airlinesLIDAR LightDetectionandRanging-Laser-based3DscanningandsensingMaaS MobilityasaService-MobilitysolutionsthatareconsumedasaserviceratherthanpurchasedasaproductODD OperationalDesignDomain-DefinitionofwhereandwhenavehicleisdesignedtooperateOEM OriginalEquipmentManufacturer-TheoriginalproducerofavehicleoritscomponentsROI ReturnonInvestment-PerformancemeasureusedtoevaluatetheefficiencyofaninvestmentSAE SocietyofAutomotiveEngineers–USbasedprofessionalassociationandstandardsdevelopingorganizationUAV UnmannedAerialVehicle-AnaircraftpilotedbyremotecontroloronboardcomputersV2V VehicletoVehicle–Wirelessexchangeofdatabetweennearbyvehicles

V2X VehicletoExternalEnvironment-Wirelessexchangebetweenavehicleanditssurroundings

Text©FutureAgenda2018Images©istockimages.com/CorporateimagelibrariesGraphs © As referenced

FirstpublishedSeptember2018by:FutureAgendaLimited84BrookStreetLondonW1K5EHwww.futureagenda.org

Glossary

3

Autonomous Vehicles The Emerging Landscape

An Initial Perspective

Tim JonesRichard BishopCaroline DewingMark Priest

Contents

SettingtheScene 5

Hype vs. Reality 6

Where We Have Come From 6

Controlled Environments 7

TheForwardView8

Purpose of this Document 9

CommonMisconceptions 10

TheBenefitsofAV 12

Peoplevs.Goods 15

GoodsVehicles 16

UrbanDelivery 18

PassengerVehicles 19

PublicTransport 21

Drivers of Adoption 22

PublicOpinion 23

Technology 25

Technology Readiness 27

Regulation 28

Insurance 29

CentresforInnovation 30

ManagingtheTransition 31

SeaborneAV 33

Air-basedAV 36

SummaryandKeyQuestions 41

References 42

ContactDetails 45

5

Imagineaworldwhereprettymucheverythingmovessmoothly,efficientlyandsecurely.Peopleandgoodsarrivewhenandwheretheyneedtobewhilemakingefficientuseofavailableresources.Thisisafuture seen through the lens of the widespread adoption of autonomous vehicles.

We are on the cusp of the transformation of the automotiveindustry.SomuchsothatMaryBarra,CEOofGM,suggeststhatwewillseemorechangeinthenext5to10yearsthaninthepast50.Thesector,whichcurrentlyhasrevenuesofover$2tnperannum,isexpectedtorepositionitsfocusfromproduct sales to service delivery and in so doing revolutionisethewaypeople,goodsandservicesmoveabout.Thisissuchasignificantshiftthatsomeseethatautonomousvehicles(AV)willactasa ‘catalysing technology’ with far reaching social andeconomicconsequences.Muchfocusisonland-basedAVbutthereisalsogrowingexcitementfor the sea and air.

Manyanalystsareworkinghardtoquantifytheopportunity:• BCGseesthatby2030theshiftto‘shared, autonomous and electric vehicles’ will account for 25%ofallUSjourneymiles.1

• McKinseyestimatesthatupto15%ofallnew vehiclessoldin2030couldbefullyautonomous.2 • Accenturesuggestsby2035asmanyas23mAVs willbeontheUShighways–justunder10%ofall registered cars and trucks.3

• GoldmanSachshasforecasttheglobalAV markettobe$96bnby2025andthatby2050the totalannualeconomicbenefitofAVadoption couldbeover$3.5tn.4

While there is growing enthusiasm and investment underway,manyrecognisethat,independentoftechnologyavailability,itisgoingtotakesometimetochangethewholevehiclefleet–maybeupto25or30years.Therearejustover1.3bnvehiclesintheworldtodayandaround100mnewonesaresold every year - so simple replacement with no marketgrowthwouldtakeatleast13years.Addinaprojectedadditionofanother700mvehiclesoverthetwodecadesand,fromlaunch,somearetalkingmorethan20yearsastheminimumforsignificantchangeinthetotalfleet.5Others consider that it may bequickerasperhapswehavealreadyreachedpeak car in the US and Europe. BCG suggests that by2030globalsaleswillplateauataround100mannuallyandthat,by2035,30%ofthevehiclefleetwillbeelectricand25%willbeautonomous.6 AccordingtoKPMG,by2030,75%oftheUKmotor-park(vehiclesinuse)willcompriseconnectedvehiclesandaround40%willbepartiallyautomated,butlessthat10%willbefullyautonomous.TheUKGovernmenthasanobjectivetoseefullydriverlesscarsonpublicroadsby2021.

Recent assessment suggests that initial adoption rateswillbefasterinEuropeandtheUS(20%by2025)thaninAsia(10%by2025)butdeeper in Asia lateron(75%by2035)thanEuropeandUS(30%by2035).7 A mandate from the Chinese central governmentisthat50%ofallnewvehiclessoldinChinaby2020musthavepartialorfullautonomousfunctions.8Globally,thereareclearlygreatexpectationsfromAV,buthowwillthispotentialchange actually occur? If you compare to other transportinnovations,automatictransmissiontook50yrstoscale,GPStook35yrswhileairbagstook25yrs.9WillAVbefaster?Willitbeslower?

What is a realistic view of market penetration for AV by 2030?

Independentoftechnologyavailability,it is going to take some time to change thewholevehiclefleet–maybeupto25or30years.

Setting the Scene

6

Twodecadesago,in1997theUSNationalAutomated Highway System Consortium demonstrated eight cars operating a single platoon ona12kmtesttrackinSanDiego.Thiswasaseminal moment in AV innovation. Since then technologyhasbeendevelopingapace.Spurredonbyinitiativessuchasthe2005DARPAGrandChallenge,therehavebeenmanynewproof-of-concept developments from a range of organisations includingbigtechcompanieslikeGoogleaswellasmajor truck manufacturers such as Scania and Volvo.

Everyoneitseemsnowwantsabitoftheaction.OrganisationslikeWaymoandTeslahavebothcapturedmediaattentionand,significantly,highprofilebutmoresecretivecompaniessuchasAppleandAmazonarealsoworkingintheAVspaceasaresomelesswell-knownbuthighlyinnovativefirmssuchasNvidia,MobileyeandNXP.Automanufacturersarealsokeennottobeleftout-backin2008GMannounceditwouldstarttestingdriverlesscarsin2015aheadoflaunchingin2018.FordandVolvoquicklysharedsimilarambitionsandtoday virtually every OEM has a self-driving division.

Acrosstheclaimsandassumptionsbeingmade,what is hype and what is reality? Evangelists believethatashifttoAVwilleliminatecongestion,massivelyreduceroadinjuriesanddeaths,freeupbillionsofdollars’worthoftime,optimisetransportsystems,reduceenergyconsumption,improveairquality,drivesocialinclusion,createhundredsofthousandsofnewjobsandincreaseGDP.Othershavemoreconservativeviewsandargue,forexample,thatiftheoverallambitionisforacleaner,low-carbonenvironmentthen,ratherthansupportingsubstantialbutuncertainAVinvestment,asimpleapproachcouldbetomakeiteasierforpeople to walk or cycle round cities. Some even suggestthatattheendoftheday,theautonomousshiftintransportationisreallyonlyaboutincreasing

trafficdensitywithoutincreasingcongestion.Whateveryourview,itiscertainlythecasethatAVis currently considered an investment opportunity and is enjoying its time in the media spotlight. Understandably,manywanttoknowwhatthetruepicture is and how will this all play out.

Hype vs Reality

Iftheoverallambitionisforacleaner,low-carbonenvironmentthen,ratherthansupportingsubstantialbutuncertainAVinvestment,asimpleapproachcouldbetomakeiteasierforpeople to walk or cycle round cities.

How important will international

standards and commonly shared

technologies be for AV adoption - or will it be more

regional?

Where We Have Come From

7

Awayfromthehighwaysandcities,autonomoustechnologieshavebeeninuseinanumberofoff-roadapplicationsforawhile.AVtechhasbeendevelopedandtestedinmanycontrolledenvironments.TheportofRotterdamwas,forexample,thefirstportintheworldwithautomatedguided vehicles.10Likewise,incountriessuchasAustralia,thebenefitsofrobotictractorshavebeen in the headlines for some time with trials now underway.11Inaddition,havinglaunchedin2015,12 many of the large mining vehicles in Australia are alsoself-driving,13whileVolvohasbeentestingself-drivingtruckswithinaSwedishminesince2016.14 AlsoinSweden,Einridehaslaunchedallelectricautonomous logging trucks that are designed to navigatehilly,windingforestroads.15Self-driving podshavebeeninuseatvariousairportsincludingLondon’sHeathrow,withotherautonomousvehiclesalsoplannedforGatwick,Paris’CDGandTokyo’sHaneda.

Lookingaheadmanyseeairportsastheidealtest-bedfornewAVtechnologies.Themainairlineindustrybody,IATA,proposesover40furtherusecasesforAVsatairportsincluding,forexample,baggagehandling.16 However one key point made byexpertshereisthattheseautomatedsystemsare all operating in essentially fully controlled environments.Theyareseparatedfromothertrafficand so the navigation technology is comparatively basic.WhilecontrolledenvironmentshavedemonstratedsomeearlystepsforAV,therearemorecomplexchallengestobeaddressedontheopen road.

Awayfromthehighwaysandcities,autonomous technologies have beeninuseinanumberofoff-roadapplications for a while.

Controlled Environments

Will controlled environments such as mines, freight terminals and airports be valuable proving grounds for wider AV development?

8

SeveralhavebeenmappingthefutureAVroadmap.Themostcommonlyuseddefinitionofautomationlevels is that of the Society of Automotive Engineers (SAE)whichidentifiessixseparatelevels(L0-L5)ranging from fully manual to fully automated systems. Thisclassificationsystemisbasedonthesplitofresponsibilitybetweenthehumanandthecomputersystem,fromallhumanresponsibilityatL0toallcomputerresponsibilityatL5.

The Forward View

ForAVtogetrealtractionitmaybenecessary to turn transport planning on its head.

Driver OnlyL0

L1

L2

L3

L4

L5

Assisted Driving Driver holds wheel / pedal controls

Partial Automation

Driver ready to regain control

Driver monitors at all times

CompleteAutomation

ConditionalAutomation

Vehicle steers andcontrols speed

Vehicle drives itself but not100% safely

Vehicle drives itself but maygive up control

Vehicle drives itself in specific cases(e.g. urban streets)

Driver not requiredat all times

Vehicle drives itself in all situations

SignificantAutomation

Driver operates vehicle

Source:https://www.sae.org/standards/content/j3016_201806

While widely adopted and logical in order for discussingvariousapproachestoautomation,somesuggestthatthe6levelsshouldnothoweverbeinterpretedasrepresentingasequentialdeploymentpath.Infact,somelevels(suchaslevel3inwhichahumanisrelieduponforasafetyfall-backrole)maynothaveasufficientbusinesscasefordeployment.

Havingideas,evenbuildingprototypesiscomparativelyeasy;ensuringtheybecomearealityinthewidercommunityismuchmorechallenging,particularlywhenitinvolveschangingthestatusquoand dealing with human interactions. Some suggest thatforAVtogetrealtractionitmaybenecessarytoturntransportplanningonitsheadand,ratherthan follow the traditional approach of predicting transportneedsandplanningaccordingly,itmaybenecessaryfornationsandcitiestodeliveranew

vision which in turn will drive demand. To do this the keywillbetounderstandwhatthevariedambitionsofmanufacturers,technologistsandgovernmentsare,howtheyintersectandalignandsowhatcanbedelivered.Thisiswhyaglobalratherthanlocalconversationisimportant.Uncoveringthebiggerpictureandrecognisingdifferentperspectivesfrommultiple regions and companies will provide a richer outlook that can then help guide some of the pivotal decisions that lie ahead.

Furthermore,whileitiseasytogetdistractedbycurrenttrendsandtheshort-termneeds,ifwelookahead,beyondtheimmediatetransportationproblemsandconsiderthe20to30-yearhorizonwemayseeasignificantalternativefutureinwhichtheAVambitionhasdeliveredchangeacrossmanyareas,notjustonland,butalsoonandunderthesea and as well as in the skies.

Will we see automation

equally across all levels or will we by-pass some?

What sociopolitical forces may

accelerate the adoption of full

automation?

Taxonomy for Driving Automation Systems for On-Road Motor Vehicles

9

Ahead of a multi-country open foresight programme toexploretheemergingAVlandscape,thisdocumentprovidesaninitialperspectivethatbringstogetherwhatisalreadybeingsharedanddiscussed.Itisbasedonlargelypublicinformationplusseveralexpertinterviews and is designed to capture existing thinking sothatwecanchallengecurrentassumptions,debatelikely developments in pivotal regions around the worldandbuildaninformed,collaborativeviewofAVoverthenext20to30years.Theresultingsynthesiswillbecuratedintoacomprehensivereportthatwillbesharedwidelytohelpguidepolicydevelopment,identifyinnovationpriorities,highlightpotentialindustrypartnerships and support investment decisions.

Purpose of this Document

This document provides an initial perspectivethatbringstogetherwhatisalreadybeingsharedanddiscussed.

10

Beforeexploringtheideasthatarebeingproposed,therearetwocritical issues to highlight that are increasing sources of confusion for AV dialogue.Thesearethedifferencebetweenconnectedandautonomousvehiclesandwherepotentialsafetybenefitsmaycomefrom.

Connected Vehicles vs. Autonomous Vehicles

Firstly,distinctionneedstobemadebetweentheincreasingly wide range of connected vehicles across the world’s transport systems and the lessdeveloped,butpotentiallymoreimpactful,autonomous vehicles of the future. Connected vehiclesarenotnecessarilyautonomous,norareAVsinevitablyconnected.Anautonomousvehicleis one that can fully operate and guide itself without human interaction whereas a connected car is one that shares information with other systems via wireless networks. Most new vehicles are already connectedtosomedegreeandmanysignificantinnovationswhichdonotrequireautonomyarealready in the pipeline.

Thereisclearlyoverlap-forexample,improved AVtrafficflowswilldependonenhanced vehicle-to-vehicle(V2V)connectivity.Connectivityand autonomy are certainly complementary technologiesandtheend-pointmaywellbethattherearevehicleswhicharebothconnectedandautonomous.Indeed,anumberoforganisationsincludethetermCAVs(ConnectedandAutonomousVehicles)inmuchoftheirstrategydevelopment.SeveralleadersinthefieldarehoweverconcernedabouttheunintendedconsequencesofthefocusonthisCAVcombinationandarguethat,asmanyofthebenefitsfromconnectivitywilloccurirrespectiveofadvancesmadeinAV,implicitlylinkingtogetherthe two areas can unnecessarily integrate decisions that may have greater impact if taken separately.

Impact on Safety

The second potential area of misunderstanding is theimpactthatcanbedeliveredforsafety.Manyareunderstandablyengagedonandbythebenefitstobegainedfromnewtechnologyimprovingthesafetyoftransportation–whetheronland,seaorintheair.In2015therewere1.25mdeathsontheroadsworldwideandover40,000fatalitiesintheUSalone.17Globallyroadtrafficinjuryisthe#1causeofdeathamongpeopleaged15to30.18In many of the world’stransportsystemsthe#1priorityisalwayssafety,andthisdrivesmostpolicydecisionsbothatlocalandnationallevels.Safervehiclesandbetterinfrastructure are clearly needed – and progress towardsthisambitionisvariouslyunderwayacrossmany regions and sectors.

Common Misconceptions

Connected vehicles are not necessarily autonomous,norareAVsinevitablyconnected.

In2015therewere1.25mdeathsontheroadsworldwideandover40,000fatalitiesintheUSalone.Globallyroadtrafficinjuryisthe#1causeofdeathamongpeopleaged15to30.

Will a change in the vehicle fleet from AV

accelerate safer transportation?

11

Intheautomotiveindustry,afrequentassertionisthatAVwillreduceaccidentsby90%.NotonlydoestheOECDseethatthisclaimisuntested,19 butothershighlightthatmostofthismaybegainedfromothersources.Forinstance,ADAS(advanceddriver-assistancesystems)havealreadymadeahugeimpact.Asoneexample,since2002autonomousemergencybraking(AEB)hasbeenfittedtoVolvoXC90swhichcorrespondinglyhaveazero-fatalitysafetyrecord.Asitbecomesstandardinmorevehicles,ThatchamResearchestimatesacrosstheUKaloneAEBcouldsaveover1,000livesandpreventover120,000casualtiesoverthenext10years.20Other ADAS innovations include adaptivecruise-control,drivermonitoring,forwardcollision warning and lane departure warnings.

Proponents are attaching many of the safety benefitsfromADASandothersimilardevelopmentstoAV.Forexample,aUKindustryviewisthat,by2030,AVwillsave2,500livesandprevent25,000accidents.21Doesthisincludethe1,000livestobesavedbyAEB?IntheUS,BCGsuggeststhatoverthesametimescale,shared,autonomous,electricvehicleswillmean25,000lessdeathsand3mfewer accidents.22 There are certainly some distinct advantages that fully autonomous cars can provide –especiallyovercominghumanbehaviourssuchasdrunkandrecklessdrivingbutthesearerelativelysmall. Virtually everyone advocating AV claim huge safetybenefits,butthemajorityofthesebenefitsmay well come from ADAS. Many highlight that somesafetybenefitsareincorrectly‘mashed’intoAVanalysis:“IfAVdoesn’thappenforsomereason,ADASwill,andsocrashrateswillgodownanyway.”TheyseethatAVisnot,initself,responsibleforsafer roads and less accidents. Rather it is part of the mix.

Virtually everyone advocating AV claim hugesafetybenefits,butthemajorityofthesebenefitsmaywellcomefromADAS.

What are the distinct benefits from AV that are not already coming from current and future-connected ADAS?

12

The Benefits of AV

Excluding safety, is the ROI on other benefits from

AV compelling enough for the

investment?

Irrespectiveofsuchconfusion,momentumbehindAVadoptioninpassengerandgoodsvehiclesisbuildingglobally.FormanythereisgrowingagreementthatAVswillatsomepointbesufficientlysafe,convenientandaffordabletodisplacemosthuman-operatedvehicles;willprovideindependentmobilityto non-drivers; reduce the stress and tedium of driving; andmaybebecomeapanaceaforcongestion,accidentandpollutionproblems.Allofthis,somesay,by2030.23Otherssuggestitismorelikelytobeby2050.However,whilethereiswidespreadalignmentonsomepotentialbenefits,currentopinionsdiffersignificantlyonthesocio-economics.

Intermsofthebignumbers,GoldmanSachsanalysissuggeststhattheglobaleconomicbenefitsfromautonomousvehiclescouldamountto$3.5tnayearby2050.24Although much of this is accounted forby$1.2tnfromtheestimated90%reductionintrafficaccidents,therestincludes$1.3tnattributedtoincreasesinmobilityand$0.9tnforproductivity.Geographicallythisisspiltas$820bnforNorthAmerica;$900bninEuropeand$1.3tninAsia.

Othersbelievethatby2030thegainsfromreclaimeddrivinghoursalonewillboostglobalGDPby$1trillion.25Much of the non-safety savings come from a reduction in travel time due to assumed significantdropsincongestionandthecostoftravel. Based on the supposition that the typical AV car operating with a shared vehicle system will coverasmanyas100,000milesperyear,BCGresearchpredictsthatby2030thecostoftravelwillhavereducedbyasmuchas50%andthatby2050AVswillgenerate30bnhoursof‘reclaimedtime’currently wasted in driving.2

But,alongwithsavings,AValsopresentsopportunitiesforgrowth.Indeed,EUanalysisproposesthatitwillcontribute$17trilliontotheEuropeaneconomyby2050andbyassoonas2020AVwilladd0.15%theannualGDPgrowthrate.27 AUKindustryviewsuggeststhatby2030AVcarswillcontributeover£50bnandcreate320,000newjobs.28BycontrasttheUKgovernmentveersonthesideofcaution,claimingthatjobcreationwillbenetpositiveand$10bnnewGVAwillbeaddedtoUKeconomyby2035.29

Some also suggest that AVs will have particular impact in economically deprived areas. SAFE (SecuringAmerica’sFutureEnergy)estimatesthecumulativebenefitsfortheUSalonewillhavereached$6.3bnby2050,includingadoublingofthetypical10mile‘reasonablecommute’fromtodaywhilesimultaneouslyenablingthe‘driver’toloseless income from dead time.30Others see that many ofthebenefitsfromAVmaycomefromsecondandthird order impacts that the mainstream are not yet anticipating.

The typical AV car operating with a shared vehicle system will cover as manyas100,000milesperyear.

Theglobaleconomicbenefitsfromautonomous vehicles could amount to $3.5tnayearby2050.

13

Akeyquestionhereiswhowillbethewinnersinthe race for AV adoption? The widely shared mantra isthatmobilityisnolongeraboutproduct,butitisaservice.Therefore,wecannolongerassessthe‘mobility’marketbyonlylookingatvehiclesales.With this in mind BCG is warning that value creation is shifting away from OEMS and towards the new tech suppliers.31Inparticular,profitpoolsaremovingtowardscomponents,dataservices,mobilityservicesand connectivity services. This is now driving OEM acquisitionsandnewcorporateventures,especiallyinthecombinationofAVandEV.

ThesearejustsomeofthemultipleshiftsbeingattributedtoAVdeployment.Takingaworldwideview,acentralglobalscenarioanalysisundertakenbytheTransportSystemsCatapultintheUKenvisagesthatAVwillaccountfor25%ofvehiclesalesby2035,withcarsandvanscontributingaround95%ofthis.32However,thereisconsiderableuncertainty. The same analysis has a high growth caseforAVaccountingfor84%ofglobalvehiclesalesandalowcaseofonly8%-quitesomespread.Withinthis,totalglobalmarketvalueforthecentralcasescenarioquadruplesfrom£5bnin2020to£20bnin2025andpasses£60bnby2035.Underlying Goldman Sachs research suggests that theshareofoverallaveragevaluewillvaryforL3andL4/5vehiclesbetween£1250and£2700 per vehicle.33

AUKindustryviewsuggeststhatby2030AVcarswillcontributeover£50bnandcreate320,000newjobs.

How significant will public funding be to the success of AV deployment?

Will AV need to be aligned with EV for greatest impact?

Globaleconomicbenefitsfromautonomousdriving($bn)

Source:GoldmanSachsGlobalInvestmentResearch

$249

$261

$81

$348

$439

$74

$1,202bn

$7

$10

$4

$13

$15

$4

$47bn

$195

$199

$67

$262

$321

$72

$921bn

$331

$352

$90

$282

$499

$114

$1,337bn

$782

$821

$242

$904

$1,275

$264

$3,506bn

Accident Reduction

Congestion Reduction

Increased Productivity

Additional Drivers

Total Benefits

United States

NorthAmerica

South America

Europe

AsiaPacific

Middle East/Africa

Global

14

Beneath this enthusiasm and future planning thereare,however,voicesofcaution-particularlyaroundtheimpactthatfullautonomywillbringinreducingcongestion,wheremuchmaypotentiallyagainbeachievedbyearlierdevelopments.Forinstance,aformofconnectedautomationknown as CACC (CooperativeAdaptiveCruiseControl)illustratesthepromiseofthemarriageofautomated control with connectivity. UC Berkeley simulationshaveshownthat,ifallvehiclesadoptCACC,UShighwaythroughputcapacityofcarswilldoublefromtoday’saverageof2000vehiclesperlaneperhour.Furthermore,thebenefitsstarttoshowwhenonly30%ofvehiclesarefittedwithCACC.34Given that hundreds of thousands of cars alreadyhaveACC,atransitiontoCACCcanbeexpected pretty soon.

A move to AV will also present a peculiar set of challenges in itself. While several experts have morepositiveviews,accordingtoKMPG, three-quartersofautomotiveexecutivesthinkthatmixing autonomous and conventional vehicles will lead to severe safety issues.35Recognising that many people like to drive and enjoy their personal space,somealsoseepublicinterestinasignificantriseinsharingvehiclesmaynotbeashighasexpected.Inaddition,theremaybeapush-backagainstsignificantjoblossesfortaxiandtruckdrivers.Maybe“optimismandconfirmationbiasesriskplaguingtheprotagonistsofAVfutures.”

Value creation is shifting away from OEMS and towards the new tech suppliers.

Are the implicit job losses from

full AV acceptable to society?

Will AV increase or decrease total

traffic flow and congestion?

15

Inconsideringwheretheremaybegreatestinitialimpact,discussionsoverthepast20yearshaveprovidednotabledebatearoundwhetherAVwillbeforpeopleorforgoodsorboth.Whiledevelopmentssuch as driverless trains and other tracked vehicles helpedsetsomeinitialexpectations,forroad-basedtransportthedebatehasfocusedonwherescalewilloccurfirst.Willitbeinmovinggoodslong-distancesonhighways,willitbesmallurbandeliveryvehicles,orwillitbeinreplacingtraditionalcars? The repeated view from expert dialogue in 2004,2007and2010wasthat,whilecarswillgrabmuchofthespotlight,itcouldbethatgrowthwillfirstoccuraroundplatooned,articulatedtrucksonthe highways and small electric delivery pods within cities.Oncethetechnologyisproveninthiscontext,withminimalhumaninteractions,itthencanbeapplied to the more sensitive challenge of moving people.

However,inthepastfewyears,themoodmusichaschanged.TheadventofUber,Lyft,Waymo,DiDiandothershasshiftedperspectivessothat,whenconsideringthetransportationofpeople,thebusinesscaseisnolongeraboutAVreplacingtheexistingcarsownershipmodel,butratherthatAVwillbecomepivotalinthedevelopmentoffleetsofsharedvehicles. The ‘access not ownership’ trend has major relevance to how many now view autonomous mobility.Giventhis,itmaywellbethattheuseofAV to move people could scale in parallel with the movement of goods and not follow in their wake.

People vs. Goods

The use of AV to move people could scale in parallel with the movement of goods and not follow in their wake.

Discussionsoverthepast20yearshaveprovidednotabledebatearoundwhetherAVwillbeforpeopleorforgoodsorboth.

Will industry collaborate to select a single system for AV globally?

Will AV have greater short-term impact in moving goods rather than people?

16

Goods Vehicles

Fortrucks,highutilisationisakeypartofexistingbusinessmodels–long-distancesand/ornearconstantavailabilityarebothcommoncharacteristicsformanyoperators.Inaddition,comparedtocars,where there are numerous organisations in the mix andhighlevelsofcompetitionbetweenestablishedsuppliers,withintheworldofHGVtruckstherearefewermanufacturers,greatercollaboration,moreregularmaintenanceandastrongbusinessorientationonROI(returnoninvestment).Moreover,ifROIisachieved,thenthereislesscostsensitivity.As such the industry is already seeing several levels of transformation moving forward simultaneously duetothestrengthofthebusinesscasesforlevel1truckplatooning,level4driverlesstruckingdriverlessvehiclesoff-roadandautonomousdeliveryon city streets.

Platooning

After20yearsofplanningandtechnologymaturation,2018sawthefirstcommercialtrialsofplatooningwherealeadtruckhasanumberoffollowertrucksincloseproximity.Platooningcanbedoneatanylevel of automation. Currently logistics companies are lininguptodeploylevel1platooningthatprovidessignificantfueleconomybenefitswhiledriversremainengagedinthedrivingtask.PioneeredbymanufacturerssuchasScania,Volvo,MANandFreightliner as well as start-up companies such as PelotonTechnology,wirelesslinksbetweenvehiclesare providing constant communication and so near-immediateaccelerationandbrakingwhenneeded.A series of trucks can therefore follow each other in verycloseproximitytherebysavingfuel.Inordertoease the integration of this operational mode into publicroads,initialplatooningconsistsofjusttwotrucks,butplatoonlengthsarelikelytoincreaseincertain(especiallyremote)areasandfortherighttypes of freight.

Thereisnorealoppositiontolevel1automationatthepoliticallevelandregulatorsare,ingeneral,supportiveofplatooningsinceitofferssocietalaswellasbusinessbenefits,withoutrequiringhighlevelsofautomationandrisk.InEurope,theUK,Finland,Sweden,GermanyandtheNetherlands,have taken steps to permit the testing and deployment of platooning – the same is the case in Australia.IntheUS,17statescurrentlyfullyallowthe commercial operation of platooning and more are expected to follow.

Fully Autonomous Stand-Alone Trucks

Ultimatelythesignificantautomation(level4)ofhighway trucks is of huge commercial interest to the freight community. The eventual aim is to enabletruckstomovewithoutadriveronboardany vehicle for hundreds of miles. This will transform long-haul journeys. Coast-to-coast across the USA currentlytakes5daysduetorequireddriverbreaks.Driverlesstruckscouldachievethesamein48hours. The associated economic gains for haulage firmsfromremovingdriversforlong-distancetrucking are massive.

2018sawthefirstcommercialtrialsof platooning where a lead truck has anumberoffollowertrucksincloseproximity.

Regulatorsare,ingeneral,supportiveofplatooningsinceitofferssocietalaswellasbusinessbenefits,withoutrequiringhighlevelsofautomation and risk.

Given the conservative nature of the

trucking industry, can it embrace

automation as quickly as

advocates foresee?

Which regions will lead the

way in agreeing regulation and

setting operating boundaries?

17

When will harmonization across jurisdictions enable system-wide integration at international levels?

Intermsofinteractionwithexistingsystems,driverless trucks may interface with human-driven trucksat‘transferhubs.’Driverswillbringloadsfromlogisticscentrestothehub,adriverlesstractorattachestothetrailerandbeginsthelongmotorwayrun.Attheexitpoint,thereverseoccurs.ThisconceptwaspioneeredbyOttoTrucks-acquiredbyUberin2016.Sincethen,Waymohasannounceditsdevelopmentofdriverlesstrucking,alongwithstart-upsEmbarkTruckingandTuSimple(foundedinChinaandnowwithparalleloperationsintheUS).Regulatoryissuesarehowevermuchthornierforlevel4thanforlowerlevelautomationbecauseofthecompleteabsenceofhumansupervision,evenfromafar.IntheUSnewlegislationmayberequiredthatcouldtakeyearstoagree,while other countries such as Australia are seeking toclearthewayexpeditiously,possiblyby2020.

Coast-to-coast across the USA currentlytakes5daysduetorequireddriverbreaks.Driverlesstruckscouldachievethesamein48hours.

18

Urban Delivery

Will small delivery robots be pivotal in building public

confidence in AV?

Awayfromtheworldoflargevehicles,HGVsandlong-distancehaulage,anotherareawherethelogisticssectorisanticipatingnotablechangeisthatofurbandelivery.Thereissignificantfocusonthehealth,energyandefficiencybenefitsoffleetsofelectric,autonomousvehiclesoperatingwithinourtowns and cities.36

Improvingtheinefficient‘firstmile’and‘lastmile’haslongbeenseenasamajoropportunityforinnovationand today many companies are seeking to gain from addressing this challenge.37Forexample,DHL,theworld’slargestdeliverycompany,isupgradingitsfleetof3,400electricdeliveryvehicleswiththelatestautonomous technology. By changing its existing fleet,DHLsaysitcanuseitscurrentnetworkstoimproveefficiencyandsoenablea24/7serviceforits consumers.38IntheUK,theGatewayprojectin Greenwich is testing self-driving food delivery in residential areas.39

Elsewhere,StarshipTechnologies’six-wheeledrobotsforpavementdeliveryhavebeenpubliclytestingin20countriessince2015andarenowincommercialuseinMountainView,California.40 SimilarproductsfromthelikesofThyssenkrupp,Nuro,Marble,andRobbyTechnologiesarealsoscaling–someinpartnershipwithestablishedmanufacturers.41Additionally,robo-taxiplayerssuchasUberandWaymoprimarilyfocusedonpassengertransportalsoseeadualopportunitytoalsoofferlocal parcel and food delivery services.

Starship Technologies’ six-wheeled robotsforpavementdeliveryhavebeenpubliclytestingin20countriessince2015.

FormanyitisAmazonthatmayhavethegreatestimpactonAVadoptionforlogistics.Thecompany’srobotsarealreadywidelyusedacrossitshugewarehousesaroundtheworld.42SinceitsacquisitionofKiva,Amazonhasscaledupitsautomationcapabilityreachingover100,000warehouserobotsin2017.43Thecompanyhasfiledmultiplepatentsforautonomousdeliveryfromairshipsanddronesaswellasgroundvehiclerobots.44In2018Amazonannounced‘postman’robotsthatcanopendoorsanddeliverparcels24hoursaday.45Asthisgiantcompanywithsubstantialinternationalreachandverydeeppocketsrollsoutautonomousrobotsincities,allintegratedintoitsexpandingfootprint,manyanticipatethatitmaysettheglobalstandardforfast,low-cost,automateddelivery.46

19

Intermsofmovingpeople,twofundamentalchangesareunderwaythatareleadingustowardsshared,autonomousvehiclesbecomingthenewmodelforpassengervehicles.Asnewtechnologyenablestheshiftfromdrivercontroltotheself-drivingideal,weare simultaneously seeing the long-anticipated move from ownership to access. The rise in popularity of car leasing and the increase of short-term rental from thelikesofZipcar,Car2GoandDriveNow,suggesttomanythatthetargetfutureAVbusinessmodelisthatofprovidingaccesstovehiclesnotownedbydriversor their employers.

Mobilityissettobedeliveredasaservicewhichwillnotonlyincreasevehicleutilisationbutcorrespondingly reduce the cost-per-mile. Whereas currently privately-owned passenger cars are parked forover90%ofthedayandhavelowutilisation,fleetsofautomatedvehicles,constantlyavailableondemand,couldbeinoperationnearly24/7.Inpreparationforthisscenario,GMrecentlyannounceda platform to allow their customers to share their privatecars,47whilein2019VolkswagenislaunchingWeShare–anEVfleetsharingsub-brandforcities.48

Asisoftenthecaseformanynewhigh-techfeatures,changetakesplacefirstatthepremium-endofthemarket. Here some expect existing car owners to simply switch to an AV – when attractive. Products arealreadyavailablefromTeslaandtheCadillacCT6isthefirstvehicletoofferatruelevel2systemwhichallows‘handsoff,feetoff’drivingaslongasthedriverispayingattention.However,evenasavailabilityincreases,someforecastsproposethatpersonalAVs will continue to cost more than human-operated vehicles,butaccesstoasharedAVwillbecheaperthanhuman-operated,ride-hailingandtaxiservices.Giventhis,leadingOEMssuchasGM,Toyota,Mercedes,BMWandVWhaveallnowindicatedthatratherthansellingorleasingAVtoindividuals,thefirstproducttheywillofferwillbe‘robo-taxi’services.49 In2018,Waymoannounceditsplanstoinitiatetherobo-taxierawithpotentialtie-upswithWal-Martandothers,sheddingfurtherlightonhowtheserviceplatform may evolve.50

Passenger Vehicles

How quickly will OEMs actually introduce highly automated mass-market vehicles?

How may the slower paced introduction of AV on mass-market cars (go anywhere) interact with faster pace launch of robo-taxi services (limited geographic extent)?

Mobilityissettobedeliveredasaservice which will not only increase vehicleutilisationbutcorrespondinglyreduce the cost-per-mile.

Personal AVs will continue to cost morethanhuman-operatedvehicles,butaccesstoasharedAVwillbecheaperthanhuman-operated, ride-hailing and taxi services.

20

Duringthis‘fleet’phasetheabsolutecostofdevelopmentwillnotbeasimportantasROI.Theaim is to give the system a chance to mature and help production costs to come down over successive vehicle generations. This in turn will open the way for the mass market introduction of highly automated vehicles.Ifsuccessful,USautomatedmobilitycostsmaydeclinefrom85¢permilein2018toabout35¢permileby2035.51 Therefore many initial deployments focus on city streets where speed (andthereforerisk)islower.Thenextevolutionofthe service is expected to extend operations to high speed limited access motorways. In areas such as Silicon Valley where freeways are a vital part of the transportnetwork,thiswillbepivotaltoensuringrobo-taxiservicesreallydoservetheneedsofthelocal populace.

However,noteverythingisquitegoingtoplan,especiallyregardingthe100%safetyambition.Accidentshavecausedsometodoubtthetechnologyreadiness-despitethefactthathigh-profilecrashesinvolvingTeslaandUbervehiclesweretheresultofhuman failures rather than those of the machine. Theyarenotgoodbellwethers:Thesesortsofeventsraiseconcernthatmaybethetechnologyisnotyetasadvancedassome,particularearlyadopters,wouldlike.Itisdifficultforoutsiderstoreallyunderstandthe current development status. The real answers lie deepwithinthelaboratoriesofthemajortechnologydevelopersthatareunderstandablyreluctanttosharethe strengths and weaknesses of their respective systems.

What we do know is that there are numerous assertionsofnear-termdeploymentplansintherobo-taxi space plus a host of cities that are enthusiastic aboutservingasinitialdeploymentsites.FleetisincreasinglyseenasthewayforwardforbothtrucksandpassengervehicleswiththelikesofUber,Lyftand DiDi all moving ahead. Proponents also say that comingyearswillseethepublicdirectlyexperiencingautomatedmobilityataverylowcost.Thiscouldpossiblychangetravelpatternsandcarownershipdecisionsintheregionswhereitisavailable.Equallythere are several within the AV tech community who doubtthetechnologicalreadinessandexpectthetimelines to stretch.

What level of safety (crashes)

is acceptable for the full launch

of AV in the next decade?

Will robo-taxis be an interim step or

a key part of the future of mobility?

How rapidly will localised

robo-taxi services expand to serve

larger regions and stimulate wider social change?

Many initial deployments focus on city streetswherespeed(andthereforerisk)islower.

Maybethetechnologyisnotyetasadvancedassome,particularearlyadopters,wouldlike.

21

Public Transport

Fortheautomatedmass-movementofpeople,driverlesstrainsandtramshavebeencommonplaceforsometime.Nowthefocusisonbuses.Todayprettymucheverymajorurbanbodyisputtingautonomouspublictransportplansontheagendaand,assumingtheearlypilotsaresuccessful,many are anticipating a wider roll-out over the next decade. Trials are underway in multiple locations includingCambridge,Michigan,52Stockholm,53Perth,54FrankfurtandMunich,55 while Mercedes-Benz’sFutureBusisbeingusedinTheNetherlandsona20kmroutebetweenSchipholairportandthetown of Haarlem.56InChina,BaiduislaunchingautonomousbusesintoBeijing,Guangdong,ShanghaiandShenzhen57while in Singapore residents in three new towns can expect to ride in anautonomousbusfrom2022.58

Quitehowautonomousbuseswillworkatscaleis not yet clear as the operational cost at times of lowusemay,insomelocations,beunacceptable– just as is the case with some conventional pubictransporttodaywhereasignificantcostisin paying the human driver. Planners are therefore considering whether the operational design of autonomousbuseswillcoverallorsomeofthecurrentserviceareasandwhetherrobo-taxiscouldbeusedtoextendreachandfilltransportationgaps.While tomorrow’s driverless services may have thepotentialtoenable24/7mobilityandimproveaccesstojobsforlocalcommunities,theconstraintsofthepublicpursewillstillhavelimitations.

Acentralquestionformanymayoralofficesand governments is whether or not the wider infrastructurewillbereadyforAVs.Manycitiesandstatesarefacingbudgetaryconstraintswhichhaveledtocrumblinginfrastructure,overburdenedandoutofdatepublictransitsystems,andincreasinglyhighertrafficcongestion.Whilesomeconsumers,politicians and industry leaders see autonomous vehiclesasthesalvationofcommutingwoes,poorcoordinationbetweentransitsystems,poorurbanplanning,andlackofattentiontolong-termtransitsolutionsmaydelayorpreventthebenefitsofautonomousvehiclesfromeverbeingrealised.59

Giventhis,Uberhasrecognisedthatelectricbikesmaybeanimportantpartoftheurbantransportmixfor short trips.60IntheUK,aseriesoftrialsarenowunderwaysupportedbytheCentreforConnectedand Autonomous Vehicles.61A core focus is how driverlessvehicleswillbeintegratedinareal-worldenvironmentandhow,fromamulti-modaltransportsystemsperspective,theycanalignwithbuses,trams,subwaysandtrains.

Acentralquestionformanymayoralofficesandgovernmentsiswhetherornotthewiderinfrastructurewillbeready for AVs.

Prettymucheverymajorurbanbodyisputtingautonomouspublictransportplans on the agenda.

Which dense cities will have integration challenges that slow AV adoption?

Will automated mobility services replace, reduce or extend the reach of public transport?

Which public bodies and systems can change fast enough to accelerate AV roll-out?

22

Aswithanymajorshift,AVwillhavegreatestimpactwherethereisanalignmentofproventechnology,realsocialneedandviablebusinessmodelsforinvestmentandreturn.ThinkofmobilepaymentsinAfrica,low-costdesalinationortheescalatinggrowth in take-up of solar energy. For AV in particular thereisasignificantroleforregulationandpolicythatwillbothhelpacceleratethedevelopmentandintroduction of critical technologies and also set the parameters within which associated products and servicescanbestoperate.Inaddition,sinceinmanyregionstheproposedshiftischangingthestatusquoandsorequiresconsumerstomakeaclearchoicetomovefromoneplatformtoanother,theviewsofthe users and their willingness to switch and adopt newsolutions-andforwhatbenefit-mustalsobeconsidered.So,therearethreeessentiallyprimarydriversofAVadoptiontoconsider–publicopinion,technology and regulation.

Withinthese,weneedtoappreciatewhatdevelopmentsmaybeglobalinnature-andsopotentiallybecomeinternationalindustrystandards-aswellasthosewhichmaybemoreregional.

Which will be the pivotal

organisations, cities and

governments that control adoption

rates?

There are three essentially primary drivers ofAVadoptiontoconsider–publicopinion,technologyandregulation.

Drivers of Adoption

Driver OnlyL0

L1

L2

L3

L4

L5


Partial Automation



CompleteAutomation










Technology

Drivers of AV AdoptionPublic

Opinion

Regulation

2020 2025 2030 2035

Reduced crewe with remotesupport and operation

of certain functionsReduced controlled

unmanned coastal vesselReduced controlled

unmanned ocean-going shipAutonomous unmanned

ocean-going ship

Drivers of AV adoption

23

ParamountforthepublicacceptanceofAVswillbeconvincingconsumerstousethem–whetherasowners,operatorsorcustomers.Initialomensarepositivebutinconsistent.Overallaround45%of people consider that they are likely to use fully autonomous vehicles in the future.62 For customers acrossmultiplemarkets,thecorebenefitsfromAVareseentoincludereducedstress,theabilitytorestonlongjourneys,morecomfortandgreaterflexibility- just as achieved for a fee in some regions today via Uber,DidiandLyft.Additionaladvantagesshouldalsoincludefeweraccidents,cheapermobilityaswell as lower insurance and fuel costs.

Asyet,quitewhichorganisationswillbemosttrustedtodeliverasafe,reliableandcomfortableexperienceisstillunclear,especiallywithchangingsentiment following recent crashes. US analysis byDeloittesuggeststhatoverhalfofuswillprefertotrusttraditionalcarbrands,whileaquarterarelooking towards a specialist new company such as Tesla,andafifthwouldhavemorefaithinbig-techsuchasGoogle,AppleorAmazon.63 Additional DeloitteresearchseesthathalfofUKconsumersthinkthatAVwillbeapositiveexperience.64 In Germanyapragmaticapproachmaybeneededas the majority of potential AV users there expect to see demonstrations under realistic conditions as opposedtomanufacturerguarantees,economicdata and technical explanations.65In China there arepositiveattitudestoAVwith45%ofdriversexpecting lower risk and hence lower insurance premiums for autonomous vehicles.66 Acceptance does not come without concerns

however. These include the potential for system failure,vehiclehackingandliabilityincaseofanaccident.67 IntheUK,73%areconcernedaboutthe safety of fully autonomous cars.68In addition some aspects of AV are more attractive than others. Forexample,althoughtheprincipleofdemandresponseridesharing(vehicleswithflexibleroutestopickupanddropoffpassengersatorneartheirdestinations)isappealing,sharingaconfinedspacewithcompletestrangersisnot.Just20%ofUberridesarecurrentlysharedviaUberPool,69 and in China DiDi recently suspended its Hitch carpool system.70Whethermoreconsumerswillbewillingtoshare an AV without a driver present is uncertain.

Public Opinion

Drivers of AV adoption

ThecorebenefitsfromAVareseentoincludereducedstress,theabilitytorestonlongjourneys,morecomfortandgreaterflexibility.

Whichorganisationswillbemosttrustedtodeliverasafe,reliableandcomfortableexperienceisstillunclear.

If AV is focused on sharing rides with strangers, will this have a negative impact on personal safety?

Who will customers trust more to deliver a safe, reliable and comfortable AV experience?

24

Althoughtherearemanyviews,globallypeople-focused AV is seen to have greatest appeal for those who currently do not have access to affordablemobilityorwhoareuncomfortableaboutdriving-theyoung,theoldandthedisabled.

• For the elderlyAVsoffertheabilitytomaintain flexible,personalmobilityforlonger.Asthe numberofthoseover75increasesby50%in mostG8countries,manyseethisasasignificant sector for growth.

• For the disabled the opportunity to overcome mobilityconstraintsisalsohighlyattractive. Especiallyforthosewerehavepreviouslybeen unabletodrive,andsohadtorelyoncarers,the potential for independent transport on demand mayopenmultipledoors,includingaccessto work and education.

• For the young the potential of lower cost and moreflexibletraveloptionsisalsoappealing, especially if it improves access to tertiary educationandbetterjobs.Currentlythoseliving inurbanareasaremostwillingtoconsiderusing driverless vehicles in the future.71

Will a marketplace of the old, young

and disabled justify initial AV

investment?

25

4500

4000

3500

3000

2500

2000

1500

1000

500

0L3 in 2015

Proj

ecte

d pa

ckag

ed v

alue

by

com

pone

nt (

£.20

15)

L3 in 2035 L4/5 in 2025 L4/5 in 2035Passive componentsEmbedded modemUltrasonic sensorsOdometry sensorsOther electronics & architectureActuators

HMI hardwareHMI softwareData security softwareMapping hardwareMapping softwareEmbedded controls hardware

Embedded controls softwareV2X hardwareV2X softwareCamerasRadarLIDAR

6431015

323617

309

231

295

169

145121145

162121155

338

288

592

394

394

494

263

214

184

375

250

250

312

167

Toaffectthedesiredchange,itisevidentthatnewtechnologyhastobedeveloped,testedandintegrated across a wide range of systems. These includecontrolsystems,sensinganddiagnosticsystems,cameras,LIDAR(LightDetectionandRanging),locationandmapping,connectivitysystems,localandcloud-baseddataprocessing,AIandmachinelearning,securitysystems,safetysystems,andhuman-machineinterfaces.72 Associated Goldman Sachs assessment of technologydemandsby2035(prioritisedbytotalglobalvalue)areLIDAR$80bn;radar$40bn;V2X$32bn;cameras$31bn;mapping$23bn;HMI$20bn;controls$17bn;security$15bnandactuators$10bn.73

Inaddition,withincreasinglysmarttechnologiesbeingappliedtotheinfrastructure,furtherareasofR&Dfocusinclude:74

• Sophisticatedenvironmentmonitoring,• Trafficcontrol,includingadvancedpredictive modelling,• Remote support to guide vehicle in tricky situations,• Verificationandvalidationprotocols,• External signalling to other car drivers and pedestrians of the intentions of the AV and • The potential need for real-time ‘road certification.’

TechnologyOf all the technologies in the mix, which ones are in greatest need of further development before the benefits of AV can be realised?

To what extent will second generation deep-learning and self-learning AI impact AV?

Costsofautonomypackagesbytechnology2035

Source:TransportSystemsCatapult-MarketForecastforConnectedandAutonomousVehicles2017

26

Thefulllistofpotentialtechprovidersissubstantialand rapidly evolving. There are numerous AV start-upsbeingfundedaroundtheworld.75Todate,alongsidemovesbyTesla,UberandWaymo,wehaveseenmostactivityinspin-offsofAV-specificsubsidiariesfromFord,Delphi,Volvo/Autolivandothers.TherehavealsobeenafewnotableM&Aactivities:

• In2016,GMpurchasedSanFranciscobased start-upCruiseAutomationfor$1bnin2016.76

• In2017,IntelacquiredIsraelicomputervision leaderMobileyefor$15.3bn.

• ThesameyearFordannounceda$1bn multi-yearventurewithArgoAItocombineboth companies’roboticsandartificialintelligenceexpertise.

Intermsofpatentactivity,thecommonindicatoroftechnologicalinnovationoutput,WaymoisbyfarthemostactivecompanyintheAVspacefilingmoreapplications than second and third places GM and Fordcombined.77

Driver OnlyL0

L1

L2

L3

L4

L5


Partial Automation



CompleteAutomation










Technology


Opinion

Regulation

2020 2025 2030 2035





ocean-going ship

Waym

o

GM

Ford

Honda

Hyundia

Toyota

Nissan

Daimler

VW

Volvo

2,118US

1,160944 870 809 716

620 604 535430

Self-drivingvehiclespatentsfiled2007-16

Source:LEKConsulting,USPatentOffice2017

Will OEMs develop their

own technologies or rely more on

partnerships and acquisitions of

tech venture and start-ups?

Intermsofpatentactivity,thecommonindicator of technological innovationoutput,WaymoisbyfarthemostactivecompanyintheAVspacefilingmore applications than second and thirdplacesGMandFordcombined.

27

Source:LEKConsulting,USPatentOffice2017

Technology Readiness

Withsomanypotentialtechnologiesinthemix,there are multiple competing views of which ones areclosesttomarket,whichwillhavegreatestimpactearlyon,whichwillbecometheagreedstandards for scaling and which will win out in the longerterm.Withinthis,manyarelookingnotjustattheAVmarketbutalsoatthewiderintegratedtransport system landscape and how all the future optionscanoperatewithinaflexible,multi-modalmodelinasafeandaffordablemanner.

Aswellasthetestingandcertificationofcoretechnologies,herecompaniesarealsodealingwiththechallengesofreliability,redundancy,cybersecurity,integrationofAI,verification/validation,aswellasotherfundamentaltenantsofautomated vehicle design such as monitoring and self-awareness.Fewinnovationshavebeenasimilarcatalyst for such widespread industry change.

Intermsofspecifictimeframesforscale-up,somebelievethedateformassintroductionofAVsiswhen we are certain that they won’t kill anyone. Others take the view that is it when they can start tosavelives.RANDlookedatthisindetailin2017andconcludedthatputtingAVsontheroadbeforethey’re perfect improves the technology more

quickly—andcouldsavehundredsofthousandsoflives over time.78“Waitingforthecarstoperformflawlesslyisaclearexampleoftheperfectbeingtheenemyofthegood.”

Waiting for the cars to perform flawlesslyisaclearexampleoftheperfectbeingtheenemyofthegood.

Asinaviation,vehiclesmusthavetheabilitytosensefailuresandswitchtoafail-safeorfail-operationalmode.Whenaproblemisdetected,thesystemthenseeksa‘minimalriskcondition’whichmaybeassimpleaspullingthevehicleofftothesideoftheroadandsignallingtoafleetmanagementcentrethathelpisneeded.Thismaybefineintestingbutnotfordeliveringmass-mobilityservices.Therefore,someareadoptinga‘remotesupport’paradigm,inwhichtheAVistetheredtoacloud-basedoperationcentrethatisstaffedbyhumanswhocanremotelyinterveneinunusualsituations.Forinstance,roadworksmayhavealaneblockedwithaflag-personwavingtrafficintotheopposinglanetogoaroundit.Somedevelopersseektohavesystemswithsufficientintelligencetounderstandallhumanflag-waversandtoknowthatinsomecasesitisacceptabletodriveintrafficdesignatedforopposingtraffic.Conversely,Nissanhaspioneered‘SeamlessAutonomousMobility’conceptinwhichahumanmobilityoperatorisabletoviewthesituationthroughthevehicle’ssensors,understandthesituation,andplotanewpath.79 The vehicle then usesitsautonomouscapabilitytomovechangedirectionandthenresumenormalindependentoperations.SinceNissanannouncedthisin2017,manyotherself-drivingcardevelopershaveadopted similar approaches. This is another means of managing the transition – rather than waiting for‘totalcapability’tobedeveloped,companiesaimtogotolaunchwithpracticalandscalablemethods that may involve intermittent human support.

Will regulators take a heavy-handed approach to new technologies or will they be light touch and maybe rely on self-certification?

How quickly will technology development move from assistive to partial and then full autonomy for the mass market?

How can AV technologies be adopted faster than the 25 to 30-year industry norms?

Human Remote Support

28

Where is advanced

regulation most likely to act as a

catalyst for AV deployment?

How will regulators

prioritize enabling deployment of various types

of automation for people and

goods?

In which countries will the legal

system be able to support AV?

ThethirdcoredriverofAVadoptionisregulation,andbyimplicationtheavailabilityofinsurance.Withoutthese,nomatterhowcompellingtheconsumerdemandandhowcleverthetech,products will not get onto the streets. Policy makers willhaveamajorimpactonhow,whereandwhenAVisrolledoutanditisaresponsibilitytheyaretakingveryseriously.Ina2018EUsurvey,87%indicated that action in AV was the top regulatory priority;morethanformedicalrobots,drones,humanenhancement,orhumancarerobots.80

Although several companies have focused in the earlystagesofdevelopmentonprivatetest-tracks,without government sanction the use of AV at scale onpublicroadsinmanycountriesiscurrentlyillegal.The1968ViennaConventiononRoadTraffic,whichhasbeenratifiedby74countries,stipulatesthat“ahumandrivermustalwaysremainfullyincontrolof,andresponsibleforthebehaviouroftheirvehicleintraffic.”SomemodificationstothisforAVarenowunderwaybutareyettobefinalised.However,notallnations,andspecificallyneithertheUKnortheUS,aresignatories.

Today,arangeofcountrieshaveorareintroducingregulationforself-drivingcarsandtrucksonpublicroads.81Globally,asofMarch2018,Canada,partsoftheUS,severalEuropeancountries,theUAE,Russia,China,SouthKorea,Japan,AustraliaandNewZealandhaveallopenedsomedoorstoAVontheirroads.Generally,thefocusismoreontestingthandeploymentand,giventhevarietyofbusinessandenduserfactorsinvolved,deploymentlocationsmay or may not align with testing locations.

Regulation

Globally,asofMarch2018,Canada,partsoftheUS,severalEuropeancountries,theUAE,Russia,China,SouthKorea,Japan,AustraliaandNewZealandhaveallopenedsomedoors to AV on their roads.

29

Insurance

Integratedtotheregulatoryshiftsisthequestionofhowvehiclesandpeopleareinsured.Globallytheemergingviewisthatautoinsurance,oneofthemostprofitableareasoftheinsurancesectorisfacing a major disruption.82 As was pointed out in oneofour2007discussions“ifcarsdon’tcrash,whoneedsinsurance?”Certainlyby2040manyusersofvehiclesthathaveADAS,orareindeedautonomous,willseenoneedtohavepersonalaccidentinsuranceandequallysocietymayjustmandate minimum coverage. Traditionally there has beena‘three-pillar’systemforvehicleinsuranceinmany markets ensuring that there is cover for the driver,theownerandthemanufacturer.Thisviewisnowinflux:Differentcountriesaretakingalternativeviewsandagain,intheUStheapproachchangesfrom state to state.

AnOECDstudyclaimedthattheburdenofliabilitywillremainthemostimportantbarrierforthemanufacturers and designers of autonomous vehicles:83ManyhavebeenexpectingthatintheearlydaysautomakerswillassumeliabilityandGoogle,VolvoandMercedesBenzarealreadydoingthis.Teslaistakingastepfurtherbyextendinginsurancetopurchasersofitsvehicles.However,whenfleetsareprovidingservicesthepersonalinsurancequestionschangesignificantly-withtheadvanceoftherobo-taximodelthingsmayevolve.Insomecountriesitmaybethatinitialdeploymentisself-insuredbygovernment.

Intermsofprofitability,arecentHBRarticle84 highlightedthatifupto94%ofUSaccidentsarecausedbyhumanerrorthen“asAVrollsoutthenumberandseverityofaccidentsandinsuranceclaimswilldrop,leadingtolowerpremiumsasinsurerspriceinaccordancewithrealrisk.”Inaddition,withmorefleetownerstherewillbemorecompetition and hence lower prices. The net impact isestimatedata$25bnlossofincomeforUSinsurersby2035outofa$200bnmarket.

To mitigate this ‘loss’ some see that new insurance productsmaybedevelopedforriskssuchascybersecurity,hackingandransomware;productfailureliabilityformanufacturersandsoftwaredevelopers;and insuring the smarter infrastructure.

Recent EU analysis highlights that there are six key categoriesofriskrelatingtotheliabilitiesofAVs:

1. Risksrelatingtothefailureoftheoperating softwarethatenablestheAVstofunction,2. Hardwarefailure,3. Networkfailures,4. Hackingandcybercrime5. Programmeddecisionmaking,and6. Risks relating to programming choice.85

IntheUS,Ford,GMandWaymohavesubmittedSafetyAssessmentLetterstotheDepartmentofTransport seeking to address these.

Applying current US rules to AVs will likely shift the balanceinliabilitydistributionbetweenconsumersandproducers,furtheraccentuateexistinggaps,andcouldpotentiallycontributetolegalandadministrative costs in connection with uncertainties. IntheEUtheapplicationoftheProductLiabilityDirective to autonomous vehicles will provide a degreeofprotection,butthereareanumberofissues that could potentially lead to decreased cover and increased costs for consumers; these include limitedreachandmeaningofproductliabilityaswellasrestrictedlistsofliablepersonsandevidentiaryburden.Somesuggestthat,overall,muchofthepremiums that consumers currently pay for auto insurance will move from the personal auto market tothecommercialliabilitymarket,specificallyintotheproductliabilitycategory.86

Much of the premiums that consumers currently pay for auto insurance will move from the personal automarkettothecommercialliabilitymarket,specificallyintotheproductliabilitycategory.

Will regions take different approaches to liability and insurance or will we see global norms?

Where may appropriate insurance lag technology availability and so hinder adoption?

Who can best drive alignment of legislation around common multinational standards?

How much of a threat is AV to the insurance industry?

30

Giventheabove,andespeciallyinordertofocusinvestmentattention,manyanalystsareseekingtodefinetheglobalhot-spotsforbothdevelopmentanddeployment.Therearemanydifferentviews.BainseesthatShanghai,London,Stockholmand Singapore are the cities leading in readying themselves and driving AV adoption.87 Others highlightSouthKorea,TheNetherlandsandIsrael.MITinBostonis,asever,significantandwhileCaliforniaisamajorcentrefordevelopment,thestatehascreatedacumbersomeregulatoryenvironment,soearlydeploymentsaremorelikelyinArizona,FloridaandothermorepermissiveUSstates. Tokyo is trialling self-driving taxis ahead of the Olympics.88The Chinese government has

beguntoshowstrongsupportforautonomousdrivingvehicletechnologyasitbeginstofast-trackthe sector.89The aim is for China to showcase to the world that it is a leading player in self-driving vehicles.

Centres for Innovation

The aim is for China to showcase to the world that it is a leading player in self-driving vehicles.

Major AV Innovation Hot-SpotsWhere will be the key hot-spots for AV development

and deployment?

31

ManyassumethatAVisultimatelyinevitable,sothebigquestionishowlongwillthetransitiontake?Theshortansweristhatitwillprobablytakelongerthan the current expectations suggest. Michigan MobilityTransformationCenterdirector,HueiPeng,reckonsthat“itmaybedecadesbeforeavehiclecan drive itself safely at any speed on any road in anyweather.”90 Several experts propose important areaswhereAVstrategyshouldbefocusedforbestmanaging the transition. These include operational designdomain,edgecasesanddedicatedlanes.

Operational Design Domain

Useinstanceswithpromisingbusinesscasesinspecificscenariosarealreadywithinreach-hencethemassivefundswhicharebeinginvestedto develop and launch new services. It is here that the concept of Operational Design Domain (ODD)becomescrucialtounderstandingcurrenttechnologyandbusinessmodeldevelopment.ForallAVs,techdevelopersdefineasetofconditionsunder which the vehicle will or will not operate in automated mode. This relates to factors such asroadtype,speed,weatherconditions,roadsignage/markingsandhighdefinitionmaps.Asoneexample,atypicalrobo-taxihasanODDwhichallows operations only on city streets where the road environmenthasbeenpre-mapped,withspeedrestrictionsof40mph,andintheabsenceofsevereweather.

Edge Cases

The development of AV systems includes a rigorous process intended to address all potential safety-relevantevents.Buthighwaysandroadscanbeunpredictable,andanyunexpectedorundetectedobstaclecouldbepotentiallyathreattolife.Therewillinevitablybenon-standardcasesinwhichtheAV system misinterprets a situation and makes an unsafedecision,possiblyresultinginhumanharm.Becauseofthis,thestakesforaddressing‘edgecases’areenormous.Todate,thecauseoftheveryhigh-profileAVaccidentsreportedinthemediarelates to human failure as drivers failed to respond properly to a safety-critical situation. Many studies rightlydiscussthedifficultiesinaddressingthemyriadedgecasesonpublicroadsand,asaresult,estimate the widespread adoption of AV services to bedecadesaway.OthersheraldAIastheanswertotheunpredictabilityoftheroadways.However,earlydeploymentsitesarebeingselectedonthebasisofa road grid and other ODD-relevant factors which canbesafelyandeffectivelyhandledbyexistingautomateddrivingsoftware.Thiswillenablethelaunch of an initial revenue-producing service while developers further evolve the software to operate over a wider range of conditions.

Managing the Transition

ForallAVs,techdevelopersdefinea set of conditions under which the vehicle will or will not operate in automated mode.

Thecauseoftheveryhigh-profileAVaccidents reported in the media relates to human failure.

Who will lead on integrating all the varied systems needed to enable AV to operate?

At what pace will ODD limitations broaden to cover most roads and most driving conditions?

32

InJuly2018Waymoannouncedthatithadreached8mmilesofon-roadtestingwithitsself-drivingfleetof600vehicles.91Thisisimpressivebutfarfromsufficient-andWaymoknowsit.Thecompanythereforeextractstheunusualreal-worldedge-casesituationsthattheircarsencounterandabstractthemintoahighlydetailedsimulationofaroadenvironment,applyingpermutationsofallpossiblevariancesonthesituation.Waymo’ssoftwareisthentestedagainstthesesituationsvirtually.Todate,Waymo’s8mroadtestmileshavebeenaugmentedby5bnsimulatedmiles.

InMarch2018,Waymoannouncedthattheywouldpurchase20,000JaguarLandRoverSUVsoverseveralyears.Inaddition,thecompanyisaddingsomeadditional62,000FiatChryslerminivanstothe600italreadyhasinuse.HavingtestedAVsinover25UScities,thecompanyalsolaunchesitsfirstcommercialhailingserviceinArizonathisyear.92

AV’sareasvulnerabletobeingstuckintrafficaswehumansaretoday.

Dedicated Lanes

Anotherkeyissueiswhethertrafficwillincreaseordecrease in an automated world - some studies say yes,someno:ManyseethatAV’sareasvulnerabletobeingstuckintrafficaswehumansaretoday.This is where the connected car technology can makeadifference.TheAVcommunityisintheearlystagesofconsideringhowcooperationbetweenvehicles and the surrounding infrastructure can significantlyimprovetrafficefficiency,bothatthestreet level and on highways. Implicit within this is thatAVswillnot,assomehadoriginallyproposed,have dedicated lanes. Although partitioned highwaysarepossibleinsomeregions,suchaspartsofAustraliaandtheUS,inmostplacestheyareimpractical,costlyand,eveniffundscouldberaised,wouldtakemanyyearstobuild.InsteadAVsneed to share infrastructure with other vehicles and users.Managingthiswillbeakeychallenge.

Thepotentialforland-basedAVisclearlyconsiderable,andthetargetbenefitssubstantial.Itis clear that with the rapid technology development underwaymultiplesolutionswillbecomeavailable.ThechallengeformanywillbeinensuringthatthetransitiontowardstheAVvisionsincreasinglybeingsharedwillbeassmoothandsustainedaspossible.

33

Seaborne AV

Althoughland-basedAVhasretainedthespotlight,thereareconsiderabledevelopmentsaroundsea-going autonomous vehicles. The merchant maritime industryiscurrentlygrowingatover3%p.a.andwilltop11bntonnesoftradein2018.93Over80%ofglobaltradebyvolumeandmorethan70%ofitsvalueiscarriedonboardshipsandhandledbyseaportsworldwide.AsitexploresAV,themaritimeindustry’s goal is not yet to remove humans from the decision-makingprocesscompletelybutrathertoeliminatetheneedforcrewtobeonboardvessels

atalltimes.Therearesignificantadvantagestothisnotleastintermsofcost,efficiencyandsafety.Withoutcrew,piracybecomesverydifficultand,sincethereisnoneedforcrewquarters,ventilation,heatingandsewagesystems,shipscanbecomelighterandmoreaerodynamic–therebyenablingthe carrying of more cargo.94Moreover,aswashighlightedinour2010research,autonomousshipsare also a key ingredient in the move to cleaner shipping–bettermanagingoceantransitroutesandtiming and so using less fuel.95

How can autonomous ships be made at least as safe as existing vessels?

ThePresidentofRolls-Royce’sMarinedivisionbelievesthat“autonomousshippingisthefutureofthemaritimeindustry….thesmartshipwillrevolutionizethelandscapeofshipdesignandoperations.”Rolls-Royce is indeed one of several companies leading the development of fully autonomous ocean-goingcargoshipswhichitexpectstolaunchinthe2035.96Thesewillbeprecededbylesscomplexharbourtugsandshortdistancecarferriesthatwillbetrialledwithinthenextfewyears.InitialfocusisontheBalticwheretheaimisforfullyremoteshipstooperateincoastalwatersby2020.97 Inaddition,theNorwegiansareundertakingtrialslookingatun-mannedelectricvesselsto transport freight along the country’s coast.98InDenmark,Rolls-RoyceandSvitzerhavealreadydemonstratedremotecontroltugsinCopenhagenharbour.99

34

Timeline for Autonomous Ship Deployment

With all the investment now underway from governmentsandcommercialcompaniesalike,the current industry view is that we will see remote controlledunmannedcoastedvesselsby2025,remotecontrolledocean-goingshipsby2030andfullyautonomousocean-goingshipsby2035.100

Tohelpenablethis,severalinitiativesareunderwayaroundtheworld:

• AnEUproject,ledbytheFraunhoferCentre forMaritimeLogistics,isassessingthe economic,legal,insuranceandtechnical feasibilityofoperatingunmannedmerchant vessels in the open seas.101

• A Swedish-led multi-country partnership is looking at safety implications102and similar researchisalsounderwayintheUK.103

• InChinaanallianceincludingRolls-Royce,ABS and Wartsila plans to deliver an unmanned cargoshipin2021,whileWuhanUniversityisone ofseverallookingathowbestautonomousships canoperatewithinbothcommercialandmilitary maritime sectors.104

• The Maritime and Port Authority in Singapore haskickedoffresearchwithNanyang Technological University; and

• InJapanNYKisplanningtotestautonomous operationofacontainershipin2019,105while the government is supporting an R&D project with Mitsui aiming to have autonomous ships operationalby2025.106

Withoutcrew,piracybecomesverydifficultand,sincethereisnoneedforcrewquarters,ventilation,heatingandsewagesystems,shipscanbecomelighter and more aerodynamic.

What will be the business incentive for owners and operators to

invest?

What technologies can best be

used to allow autonomous

operation mid-ocean?

Source:RollsRoyce

Driver OnlyL0

L1

L2

L3

L4

L5


Partial Automation



CompleteAutomation










Technology


Opinion

Regulation

2020 2025 2030 2035





ocean-going ship

35

Meanwhile,belowthesurface,theautonomousunderwatervehicle(AUV)marketisalsodeveloping.Leaduserstodayarethemilitary,scientistsandcontractorscheckingunderwatercables.BoeingisonecompanypartneringwiththeUSNavytocreatedronesubmarines,thelatestofwhich,EchoVoyager,isover50feetlonganddesignedtooperate at sea for months at a time with no need foranysurfaceshipstand-bysupport.107In2017Israel’s Ben-Gurion University launched the country’s firstautonomoussubmarine,the8-footlongHydroCamelII.Itisfullyautonomous,hashighlevelmanoeuvringandisabletodivealmostvertically.108 KongsbergMaritimealsohasagrowingrangeofsmall AUVs and has recently announced its intent to acquireRolls-RoyceCommercialMarinebusiness.

Althoughlesspublicthanland-basedAV,itisclearthatthebenefitsofsafetyandefficiencyarejust as important at sea. With rapidly advancing technologiesandcompellinginvestmentcases,assuming the myriad international legal issues can beaddressed,manyseethatthetrajectoryforseaborneAVsisperhapsmorepredictablethanothers.

The autonomous underwater vehicle (AUV)marketisalsodeveloping.

To what extent can land, sea and air-based AV share technology?

Will autonomous ships be legal and who is liable in the event of an accident?

36

Air-based AV

Lastly,thereisalsosignificantactivityunderwayinthesky.WhileNASAhasdemonstratedremotelycontrolled planes for several decades and most spaceexplorationhasusedmorecomplexsystems,theshifttoautonomousUAVs(unmannedaerialvehicles)has,inrecentyears,alsobeengatheringpace.MostsignificantlymilitarydroneshavecomeunderpublicscrutinyaspartiallyautomatedUAVshavebeguntoappearalongsidethemoretraditionalremotely piloted models such as the USAF Reaper andPredatorusedforreconnaissance,combatandweapondelivery.However,otherthanforactivitieslikeautomatedin-flightrefuelling,fullyautonomousUAVsarestillanemergingfield,largelydrivenbymilitaryambition.

AlthoughthelikesofElonMusk,StephenHawking,theUNandthousandsofAIdevelopershaveallurgedabanonwarfareusingautonomousweapons,there is no industry consensus on the main legal and ethical issues and so development continues. “Thelimitingfactorisnotthetechnologybutmorethepoliticalwilltoproduce,oradmittohaving,suchcapabilities.”Currentlythereisashiftfromonecrewremotely piloting a single drone to the same crew controllingavastsquadronviaasinglemothercraft.Militarylawcurrentlyrequireshumanpresenceinthedecisioncycleandthepotentialabsenceofthisispushingmultipleboundaries–notleasttheethicalones.Movingforwardkeyquestionsinclude;therelative importance of military advantage vs. collateral damage and the principle of proportionality; as well ashowmuchhumaninteractionisrequired.Somebelievethatthatallowingamachineto‘decide’tokillahumanisasteptoofar.Ontheotherhand,itcouldbearguedthatutilisingautonomousdronesisbothacceptableandpreferabletousinghumanpilots and soldiers in areas of high risk. The current NATOviewisthat“theresponsibilityforwarcannotbeoutsourced,leastofalltomachines.”However,theadventoffullyautonomousmilitarycombatdronesisnot far away.109

Militarydroneshavecomeunderpublicscrutiny as partially automated UAVs havebeguntoappearalongsidethemore traditional remotely piloted models.

37

Beyond military applications there is growing commercial interest in drones for people and goods.UPSandFedExarebothexperimentingwithdrones,butmostsignificantlyAmazon’sambitionstodeliverupto90%ofparcelsweighinglessthan2kg via drones have captured the imagination. Inanticipationofmajorgrowth,thecompanyhas already patented many innovations including autonomousairtrafficcontrol.110

Drones do not have the same level of collision avoidancesystemsasplanes–largelybecauseofweightandcost,sotheaimisforanewtypeofflightmanagementsystemwhichusesensors,digitalcertificationandwirelesscommunicationstocreateameshnetworkabletoself-manageUAVairrouting.AstheUSregulatoryenvironmentprohibitsthis,initialtrialshavetakenplaceintheUK,Israel,AustriaandFrance.Nokiaisalsoworkinginthisareaandin2016ranpilots-firstinTheNetherlandsandthenDubai.111

Amazon’sambitionstodeliverupto90%ofparcelsweighinglessthan2kg via drones have captured the imagination.

How realistic is Amazon’s ambition for drone parcel delivery?

38

Alongside moving packages there is growing interest in the development of autonomous air taxisforpeople,notablyinChina,theUS,DubaiandNewZealand.Althoughprobablyinitiallyonlyforthewealthy,astheyevolve,andcitiesbecomemorepopulated,some,suchasGoogleXfounderSebastianThrun,seethatairtaxiscouldeventuallycompetewithgroundtransportasamoreflexibleoptionforpersonalmobility.112

Air taxis could eventually compete with groundtransportasamoreflexibleoptionforpersonalmobility.

Prominent Flying Taxi Developments

• AirbuscompleteditsfirsttestflightofitssinglepersonVahanaflyingtaxiinJanuary2018and planstolaunchafleetofthesemulti-rotoraircrafttoflyfromrooftoptorooftopindensecitieswhere therearehighlevelsoftrafficcongestion;113

• InChina,dronemanufacturerEhangismoreadvancedandalreadyflyingtestflightsofupto 15kminGuangzhouandreachingspeedsofupto80mph.Passengerssimplymarktheir destinationandthedronecreatesandexecutestheflightplan;AirTaxibackedbyDaimlerandIntel isbeingtestedaspartofagovernmentambitionforautonomoustransporttoaccountfor25%ofall tripsby2030;114

• IntheUS,Uberisdevelopingits200mphElevateairtaxireadyfortestingin2020;and

• InNewZealand,Googleco-founder,LarryPage,hasdevelopedatwo-personplane/drone hybrid,theKittyHawkCora,thatcanreach110mphandflyfor50milesatupto3000ft. ThisiscurrentlyundergoingregulatoryapprovalandmaywellpropelNewZealandtotheforein certificationanddevelopment.115

Are air-taxis going to be affordable

beyond the very wealthy?

39

Movingupinscale,severalorganisationsareexploringthepossibilityofpilotlesspassengerflights.Itseemsanaturalprogressionasmanyplanesalreadyflyonautopilotandmanyairportsarealreadyequippedforaccommodatingfullyautomatedtake-offandlanding.Inadditionthereisagrowingneedbothpractically,(Boeingestimatesthatanother637,000pilotsarerequiredinthenext20years)andeconomically(accordingtoUBS,beingpilotlesscouldsavetheairlineindustryabout$35bnayear).116 As yet there is no regulation in placeforfullautomationandthereareobviousconcernsaboutthelackofahumanpresenceontheflight-deck,howevertheideaofremotepilotsonthegroundasbackupisgainingtraction.117Intheinterim,AirbushassetuparesearchfacilityinShenzhenwiththeintentionofusingmoreautomationtoreducethenumberofpilotsinaplanefrom2to1.118Thiscoulddeliverover40%ofthe$35bnsavingcalculatedbyUBS.

Again,alongsideseaborneAVs,withinternationalagreements already part of the system and a shorter listofcompaniesinvolvedthanwithland-basedAVs,several expect that we will see rapid growth in UAV over the next decade or so. Regulation and consumerconfidenceareevidentlyimportantfactorsthat will dictate the speed of adoption.

Thereareobviousconcernsabout the lack of a human presence on the flight-deck.

Who will manage the airspace?

41

Summary and Key Questions

Aswemoveforwardtoengageinaseriesofexpertdiscussionsinkeylocationsaroundtheworld,exploringthekeyuncertaintieswillbeapriority.Asworkshopsareheldinthedevelopmentcentresofinnovationandearlydeploymenthotspots,thefullrangeofAVrelatedtopicswillbeinterrogated.However,inordertogainarich,informedandcredibleview,fromtheassessmenttodatethereseemstobetwelvekeyquestionsthatshouldbeaddressedupfront.ThesearepivotaltohowthefutureAVlandscapewillemerge:

1. Wherewillbethekeyhot-spotsforAVdevelopmentanddeployment?

2. WhichsociopoliticalforcesmayacceleratetheadoptionoffullLevel4/5automation?

3. WhereisadvancedregulationmostlikelytoactasacatalystforAVdeployment?

4. Whatlevelofsafety(crashes)isacceptableforthefulllaunchofAVinthenextdecade?

5. WillAVincreaseordecreasetotaltrafficflowandcongestion?

6. Willautomatedmobilityservicesreplace,reduceorextendthereachofpublictransport?

7. Ofallthetechnologiesinthemix,whichonesareingreatestneedoffurtherdevelopmentbeforethe benefitsofAVcanberealised?

8. WhatarethedistinctbenefitsfromAVthatarenotalreadycomingfromcurrentandfuture-connectedADAS?

9. HowimportantwillinternationalstandardsandcommonlysharedtechnologiesbeforAVadoption-orwill itbemoreregional?

10.Whichwillbethepivotalorganisations,citiesandgovernmentsthatcontroladoptionrates?

11.WhowillleadonintegratingallthevariedsystemsneededtoenableAVtooperate?

12.Whowillcustomerstrustmoretodeliverasafe,reliableandcomfortableAVexperience?

Over the next six months, expert workshops in over 15 leading centres globally will seek to answer these and other questions. From this initial review it is evident that AV is on the verge of having major global impact. As we explore the growing potential for more automation across land, sea and air-based AV, we will better understand the varied ambitions of manufacturers, technologists and governments as well as how they intersect and align. The informed, global view will provide a richer outlook that can then help guide some of the pivotal decisions ahead.

42

References1 https://www.bcg.com/en-gb/publications/2017/reimagined-car-shared-autonomous-electric.aspx2 https://www.mckinsey.com/~/media/mckinsey/industries/high%20tech/our%20insights/disruptive%20trends%20 that%20will%20transform%20the%20auto%20industry/auto%202030%20report%20jan%202016.ashx3 https://hbr.org/2017/12/driverless-cars-will-change-auto-insurance-heres-how-insurers-can-adapt4 http://pg.jrj.com.cn/acc/Res/CN_RES/INVEST/2015/9/17/f70472c6-f4ad-4942-8eab-3c01f3c717a7.pdf5 https://www.weforum.org/agenda/2016/04/the-number-of-cars-worldwide-is-set-to-double-by-20406 https://www.bcg.com/d/news/11january2018-the-great-mobility-tech-race-1836657 https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/642813/15780_ TSC_Market_Forecast_for_CAV_Report_FINAL.pdf8 https://ihsmarkit.com/research-analysis/Chinese-government-drafts-policies-autonomous-vehicles.html9 https://www.vtpi.org/avip.pdf10 https://www.portofrotterdam.com/en/business-opportunities/smartest-port/cases/new-standard-in-container-terminals- and-services 11 https://www.futurefarming.com/Machinery/Articles/2017/6/Farmer-shares-his-experiences-using-a-robotic-tractor- 1614WP/12 https://www.telegraph.co.uk/technology/news/11940333/Driverless-trucks-transport-iron-ore-at-Rio-Tinto-mines-in- Australia.html 13 https://www.technologyreview.com/s/603170/mining-24-hours-a-day-with-robots/14 https://www.volvogroup.com/en-en/news/2016/sep/news-2297091.html15 https://electrek.co/2018/07/12/einride-t-log-all-electric-autonomous-logging-truck/16 https://www.futuretravelexperience.com/2018/04/autonomous-vehicles-future-proofing-the-airport-environment/17 http://www.who.int/violence_injury_prevention/road_safety_status/2015/Executive_summary_GSRRS2015.pdf?ua=118 http://www.who.int/gho/road_safety/mortality/en/19 https://www.itf-oecd.org/sites/default/files/docs/safer-roads-automated-vehicles.pdf20 https://www.thatcham.org/car-safety/automatic-braking/21 https://www.smmt.co.uk/industry-topics/technology-innovation/connected-autonomous-vehicles/22 https://www.bcg.com/en-gb/publications/2017/reimagined-car-shared-autonomous-electric.aspx23 https://www.vtpi.org/avip.pdf24 http://pg.jrj.com.cn/acc/Res/CN_RES/INVEST/2015/9/17/f70472c6-f4ad-4942-8eab-3c01f3c717a7.pdf25 https://static1.squarespace.com/static/585c3439be65942f022bbf9b/t/591a2e4be6f2e1c13df930c5/1494888038959/ RethinkX+Report_051517.pdf26 https://www.bcg.com/en-gb/publications/2017/reimagined-car-shared-autonomous-electric.aspx27 http://www.europarl.europa.eu/RegData/etudes/STUD/2018/615635/EPRS_STU(2018)615635_EN.pdf28 https://www.smmt.co.uk/industry-topics/technology-innovation/connected-autonomous-vehicles/29 ttps://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/642813/15780_TSC_ Market_Forecast_for_CAV_Report_FINAL.pdf30 https://avworkforce.secureenergy.org/wp-content/uploads/2018/06/Americas-Workforce-and-the-Self-Driving-Future_ Realizing-Productivity-Gains-and-Spurring-Economic-Growth.pdf31 https://www.bcg.com/d/news/11january2018-the-great-mobility-tech-race-18366532 https://www.gov.uk/government/publications/connected-and-autonomous-vehicles-market-forcecast33 http://pg.jrj.com.cn/acc/Res/CN_RES/INVEST/2015/9/17/f70472c6-f4ad-4942-8eab-3c01f3c717a7.pdf34 https://path.berkeley.edu/research/connected-and-automated-vehicles/cooperative-adaptive-cruise-control35 https://assets.kpmg.com/content/dam/kpmg/tr/pdf/2018/03/global-automotive-executive-survey-2018.pdf36 https://www.technologyreview.com/s/602896/fedex-bets-on-automation-as-it-prepares-to-fend-off-uber-and-amazon/

43

37 http://archive.futureagenda.org/pg/cx/view/411?view=acs38 http://www.alphr.com/cars/1007293/dhl-aims-for-an-autonomous-future-with-a-fleet-of-self-driving-delivery-vehicles39 https://gateway-project.org.uk/uk-first-autonomous-grocery-delivery-trials-in-greenwich/40 https://www.theguardian.com/cities/2018/apr/30/robot-delivery-drivers-coming-to-a-campus-near-you-starshiptechnologies 41 https://www.zdnet.com/article/this-pint-sized-autonomous-delivery-vehicle-is-coming-to-city-streets-this-year/42 https://www.ft.com/content/916b93fc-8716-11e7-8bb1-5ba57d47eff743 https://www.nytimes.com/2017/09/10/technology/amazon-robots-workers.html44 http://uk.businessinsider.com/amazons-robot-army-has-grown-by-50-2017-145 https://www.standard.co.uk/news/techandgadgets/amazons-postman-robots-that-could-unlock-doors-to-deliver- parcels-a3758376.html46 https://www.forbes.com/sites/parmyolson/2018/01/31/amazon-just-sparked-a-race-to-bring-robots-to-our- doors/#6c1af99c2b3a47 https://www.forbes.com/sites/joannmuller/2018/07/24/would-you-rent-your-car-to-a-stranger-gms-maven-hopes- so/#11cf3e85648f48 https://www.volkswagenag.com/en/news/2018/08/VW_Brand_We_Share.html49 https://www.vtpi.org/avip.pdf50 https://medium.com/waymo/growing-waymos-partnerships-in-metro-phoenix-96594145190251 https://www.vtpi.org/avip.pdf52 https://mcity.umich.edu/shuttle/53 https://www.forbes.com/sites/heatherfarmbrough/2018/01/31/ugly-but-useful-stockholm-introduces-driverless- busses/#3372e60560f454 https://www.perthnow.com.au/news/wa/passengers-give-thumbs-up-for-driverless-bus-ng-23861a4c2156a46452f1674 d036c853d55 https://www.dezeen.com/2017/10/27/germany-unveils-first-driverless-bus-bavaria-technology-transport/56 https://www.daimler.com/innovation/autonomous-driving/future-bus.html57 https://techcrunch.com/2018/07/03/baidu-just-made-its-100th-autonomous-bus-ahead-of-commercial-launch-in-china/58 https://www.straitstimes.com/singapore/transport/centre-for-self-driving-vehicles-opens-in-jurong-west59 https://www.autoindustrylawblog.com/2017/10/16/autonomous-cars-are-here-to-stay-but-are-cities-ready-for-them/60 https://www.ft.com/content/986d878a-a7c4-11e8-8ecf-a7ae1beff35b61 https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/725522/ccav- research-and-development-projects.pdf 62 https://www.smmt.co.uk/wp-content/uploads/sites/2/Connected-and-Autonomous-Vehicles-Revolutionising-Mobility-in- Society.pdf63 https://www2.deloitte.com/content/dam/insights/us/articles/3565_Race-to-autonomous-driving/DR20_The%20race%20 to%20autonomous%20driving_reprint.pdf64 https://www2.deloitte.com/content/dam/insights/us/articles/3913_Driverless-cars-in-the-UK/DUP_Development-of-self- driving-vehicles-UK.pdf65 https://www2.deloitte.com/content/dam/Deloitte/de/Documents/consumer-industrial-products/Autonomous-driving-in- Germany_PoV.pdf66 https://www.sciencedirect.com/science/article/pii/S096585641731059567 http://www.bain.de/Images/BAIN_BRIEF_An_Autonomous_Car_Roadmap_for_Suppliers.pdf68 https://www2.deloitte.com/content/dam/insights/us/articles/3913_Driverless-cars-in-the-UK/DUP_Development-of-self- driving-vehicles-UK.pdf69 https://techcrunch.com/2016/05/10/uber-says-that-20-of-its-rides-globally-are-now-on-uber-pool/

44

70 https://www.bbc.co.uk/news/world-asia-china-4531307471 http://tsctechstrategy.co.uk/wp-content/uploads/2016/04/Tech_Strategy_Brochure.pdf72 https://www.technologyreview.com/s/609674/whats-driving-autonomous-vehicles/73 https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/642813/15780_ TSC_Market_Forecast_for_CAV_Report_FINAL.pdf74 http://pg.jrj.com.cn/acc/Res/CN_RES/INVEST/2015/9/17/f70472c6-f4ad-4942-8eab-3c01f3c717a7.pdf75 https://techcrunch.com/2018/05/27/the-well-funded-startups-driven-to-own-the-autonomous-vehicle-stack/76 https://www.ft.com/content/16b55fe8-e79f-11e5-bc31-138df2ae9ee677 http://uk.businessinsider.com/us-uk-self-driving-car-regulations-2018-378 https://www.rand.org/blog/articles/2017/11/why-waiting-for-perfect-autonomous-vehicles-may-cost-lives.html79 https://www.nissan-global.com/EN/TECHNOLOGY/OVERVIEW/sam.html80 http://www.europarl.europa.eu/RegData/etudes/STUD/2018/615635/EPRS_STU(2018)615635_EN.pdf81 https://medium.com/syncedreview/global-survey-of-autonomous-vehicle-regulations-6b8608f205f982 http://pg.jrj.com.cn/acc/Res/CN_RES/INVEST/2015/9/17/f70472c6-f4ad-4942-8eab-3c01f3c717a7.pdf83 https://www.itf-oecd.org/sites/default/files/docs/safer-roads-automated-vehicles.pdf84 https://hbr.org/2017/12/driverless-cars-will-change-auto-insurance-heres-how-insurers-can-adapt85 http://pg.jrj.com.cn/acc/Res/CN_RES/INVEST/2015/9/17/f70472c6-f4ad-4942-8eab-3c01f3c717a7.pdf86 http://www.europarl.europa.eu/RegData/etudes/STUD/2018/615635/EPRS_STU(2018)615635_EN.pdf87 http://www.bain.de/Images/BAIN_BRIEF_An_Autonomous_Car_Roadmap_for_Suppliers.pdf88 https://www.theguardian.com/technology/2018/aug/28/driverless-taxi-debuts-in-tokyo-in-world-first-trial-ahead-of- olympics 89 https://ihsmarkit.com/research-analysis/Chinese-government-drafts-policies-autonomous-vehicles.html90 https://www.vtpi.org/avip.pdf91 https://www.theverge.com/2018/7/20/17595968/waymo-self-driving-cars-8-million-miles-testing92 https://www.theguardian.com/technology/2017/nov/07/google-waymo-announces-fully-autonomous-ride-hailing- service-uber-alphabet93 http://unctad.org/en/pages/PublicationWebflyer.aspx?publicationid=189094 https://www.techemergence.com/autonomous-ships-timeline/95 http://archive.futureagenda.org/pg/cx/view/374?view=acs96 https://spectrum.ieee.org/transportation/marine/forget-autonomous-cars-autonomous-ships-are-almost-here97 https://www.rolls-royce.com/~/media/Files/R/Rolls-Royce/documents/customers/marine/ship-intel/aawa- whitepaper-210616.pdf98 https://www.dnvgl.com/99 https://www.rolls-royce.com/media/press-releases/yr-2017/20-06-2017-rr-demonstrates-worlds-first-remotely-operated- commercial-vessel.aspx 100https://www.rolls-royce.com/~/media/Files/R/Rolls-Royce/documents/customers/marine/ship-intel/rr-ship-intel-aawa- 8pg.pdf101http://www.unmanned-ship.org/munin/about/102https://www.eda.europa.eu/docs/documents/SARUMS-Flyer-2012.pdf?sfvrsn=0103http://www.ukmarinealliance.co.uk/MAS104http://english.whut.edu.cn/105https://www.globalpsa.com/autonomous-ships/106http://www.mol.co.jp/en/pr/2017/17031.html107https://www.popularmechanics.com/military/navy-ships/a26836/boeing-teaming-up-to-build-autonomous-submarine/

45

108https://eandt.theiet.org/content/articles/2017/10/researchers-reveal-israel-s-first-autonomous-submarine/109https://www.nato.int/docu/review/2017/also-in-2017/autonomous-military-drones-no-longer-science-fiction/EN/index.htm110https://www.cbinsights.com/research/amazon-drone-air-traffic-control-patent/111https://www.mobileeurope.co.uk/press-wire/nokia-s-drones-take-off-at-twente-airport-in-traffic-management-test112https://www.theguardian.com/technology/2018/aug/26/sebastian-thrun-interview-costs-of-air-taxi-system-less-than- uber-google-glass-flyer-cora-kitty-hawk113https://www.theverge.com/2018/2/1/16961688/airbus-vahana-evtol-first-test-flight114https://www.theverge.com/2017/9/26/16365614/dubai-testing-uncrewed-two-person-flying-taxis-volocopter115https://www.theguardian.com/technology/2018/mar/13/kitty-hawk-autonomous-drone-flying-taxi-service-regulatory- approval-google-larry-page 116http://uk.businessinsider.com/pilotless-planes-could-be-a-35-billion-opportunity-wall-street-is-missing-2018-7117https://www.theguardian.com/business/2017/aug/07/pilotless-planes-what-you-need-to-know118https://www.independent.co.uk/news/business/news/airbus-pilotless-planes-self-flying-aircraft-passenger-flights-cto- paul-eremenko-a8068956.html

46

Contact details

To discuss this project further please get in touch

Dr Tim Jones

Programme Director

Future Agenda

[email protected]

www.futureagenda.org

+447801755054

@futureagenda

AUTONOMOUS VEHICLES THE EMERGING LANDSCAPE An Initial ... Agenda - Autonom… · motor-park...

Documents

Transcript of AUTONOMOUS VEHICLES THE EMERGING LANDSCAPE An Initial ... Agenda - Autonom… · motor-park...