2025年自動(dòng)駕駛汽車：技術(shù)演進(jìn)及未來路線圖-2025-05-自動(dòng)駕駛

Incollaborationwith

BostonConsultingGroup

AutonomousVehicles:

TimelineandRoadmapAhead

WHITEPAPERAPRIL2025

Images:GettyImages

Contents

Foreword3

Executivesummary4

Introduction5

1

Assisted,automatedandautonomouspersonalvehicles7

2

Robotaxisandroboshuttles11

3

Autonomoustrucks14

4

Anoverarchingindustryagenda18

Conclusion21

Contributors22

Endnotes24

Disclaimer

Thisdocumentispublishedbythe

WorldEconomicForumasacontributiontoaproject,insightareaorinteraction.

Thefindings,interpretationsand

conclusionsexpressedhereinarearesultofacollaborativeprocessfacilitatedand

endorsedbytheWorldEconomicForumbutwhoseresultsdonotnecessarily

representtheviewsoftheWorldEconomicForum,northeentiretyofitsMembers,

Partnersorotherstakeholders.

?2025WorldEconomicForum.Allrightsreserved.Nopartofthispublicationmaybereproducedortransmittedinanyformorbyanymeans,includingphotocopyingandrecording,orbyanyinformation

storageandretrievalsystem.

AutonomousVehicles:TimelineandRoadmapAhead2

April2025

AutonomousVehicles:TimelineandRoadmapAhead

Foreword

JeremyJurgensManagingDirector,

WorldEconomicForum

Overthousandsofyearstransporthasalways

connectedpeopleandexpandedtheiraccess

toopportunities,consistentlygrowingeconomies

andadvancingsocieties.Withinthatongoing

evolution,autonomousvehicles(AVs)represent

oneoftransport’smostanticipateddevelopments,offeringthepotentialtoimproveroadsafety,enhancelogisticsandenablenewmobilityservices.Thereare,however,significanttechnological,regulatoryand

operationalchallengestorealizingthosebenefits.AVsmustalsobecarefullyintegratedintoexistingtransportecosystemsasmixedtrafficconditionswillcreatecomplexitiesforyearstocome.

Itiscrucialthatthesector’sstakeholderscan

baseinformeddecisionsonrealisticexpectations,yetpredictionsaboutthedeploymenttimelinefor

autonomousvehicleshavetendedtobeoverly

optimistic.Whilevehicleautomationtechnologyhasadvancedconsiderably,itslarge-scaleintegration

willtakelongerthanmosthaveanticipated.This

whitepaperaimstoprovideamoregrounded

perspectiveontheadoptiontimeline,addressing

threekeyusecasesofvehicleautonomybetween2025and2035:personalvehicles,robotaxisand

autonomoustrucks.Itstrivestoanswersomeof

thekeyquestionsofpolicymakers,businessleadersandthepublicabouttheseevolvingtechnologies.

NikolausLang

ManagingDirectorandSeniorPartner,

BostonConsultingGroup

Thetimelineforadoptingtheseinnovations

haswidesocietalimplicationsbeyondtransportplanning,andaddressingthesechallenges

atanearlystageisessentialtoasuccessful

rollout.Forexample,manyworkersmaystruggle

toadapttochangingjobrequirements.Anaccuratetimelinecanhelpdecision-makersbetterprepare

workforcereskillingprogrammes.Dataprivacyandcybersecuritymustalsobeprioritized;autonomousvehiclesgathervastamountsofreal-timeand,

tosomeextent,sensitivedata,aboutwhatis

happeninginthevehicleanditssurroundings.

Equitableaccesstovehicleautomationtechnologyiscritical,too:AVdevelopmentmustenablemoreholisticandinclusivemobilitysystemsinstead

ofexacerbatingexistingtransportinequalities.

Widespreadadoptionofautonomousvehicleswillremainslow,butthedecisionsmadetodaywill

shapehowthistechnologyintegratesintosocietytomorrow.Governments,industryleadersandcivilsocietyneedtocollaboratetoensurethatsocietalneedsaremetandthatautonomousvehicles

contributetoamoreefficient,sustainableandinclusivemobilitylandscape.

AutonomousVehicles:TimelineandRoadmapAhead3

Executivesummary

Autonomousvehicles:Scalingforimpactwhileaddressingremainingchallenges.

Earlydeploymentsofautonomousvehiclesare

alreadyontheroads.However,itisbecoming

apparentthatlarge-scalerolloutwillbeslowerthanonceanticipated.Whilepreviousandevensome

currentforecastsstatethatautonomousvehicleswillbewidelyadoptedduringthe2020s,the

analysesofthiswhitepapersuggestmainstreamdeploymentwillbeslowerthanthatgiventhemanychallengesandinherenttechnological,regulatoryandeconomiccomplexities.

Despitethis,therationaleforAVadoptionremainscompelling,drivenbysubstantialpotentialbenefitsincludingenhancedsafety,improvedefficiencyandlowercosts.Thiswhitepaperprovidesarefined

forecastfordeploymentandidentifieskeyremaininggapsandactionsforacceleratingthatdeploymentsafely.Itexploresthreemainusecases:personal

vehicles,robotaxisandautonomoustrucks.Thekeyinsightsoneachoftheseareasfollows:

–Whilepersonalvehicleswillprogressively

transitiontowardhigherlevelsofautomation,L2andL2+systemswilldominatethisusecaseforthenextdecadeduetotheircost-effectivenessandregulatoryreadiness.L3adoptionwill

remainlimitedduetosafetyrisks,liability

concernsandhighcosts,andL4deploymentwillbenicheduringthistimeframe:onlyaround4%ofnewpersonalcarssoldby2035are

expectedtofeatureL4capabilities.China

isforecasttoadoptL2+andL3/L4vehicles

mostquickly,drivenbystrongconsumer

demand,aregulatorypushandanecosystemthatencouragesinnovation.(SeeBox1foranexplanationofthelevelsofautomation.)

–Robotaxishavealreadydemonstrated

technologicalfeasibility,withlarge-scale

deploymentsrunninginselectedUSand

Chinesecities.However,thehighcostsof

softwaredevelopment,infrastructureset-up

andscalingcontinuetoslowdeployment.By

2035,robotaxisarelikelytobepresentinlargenumbersacross40to80citiesglobally,mostlyinChinaandtheUnitedStates.Untilatleast2030,Europeisexpectedtoremaincautiousabouttherolloutofrobotaxis.Europeislikelytoprioritize

small,controlledpilotsandfocusonintegratingroboshuttleswithpublictransportsystems

instead.Large-scalerobotaxi(androboshuttle)deploymentswillleadtomodalshifts,affectingnotonlytaxiandtraditionalride-hailingbutalsopersonalcarandpublictransportuse.

–Autonomoustruckingpresentsastrongcase

forautonomy.Comparedtotraditionaltrucking,itintroducesanewvaluepropositionthatgoesbeyondadvantagesinefficiencyandtotalcostofownership.Severalcompanieshavestartedcommercialoperations,and2025isexpected

tobeanimportantyearforautonomoustruckingdeployments.Amongthedifferentuse-cases,

hub-to-hubtruckinghasthemostpromiseforautomation.TheUnitedStatesisexpectedto

leadadoptionforthisusecase:itisprojected

thatautonomoustruckswillaccountforupto

30%ofnewtrucksalesintheUSby2035.In

Europe,internationalbordersposechallengesforlong-haulapplications,andChina’sweakercostbenefitsmaylimitdeploymentunlesspolicyinterventionsaccelerateprogress.

Theforecastsinthiswhitepaperaimto

accountforexpecteddevelopments.However,technologicalbreakthroughs,suchasthe

successfuldeploymentofmap-freeandvision-

onlyL3/L4systems,ormassiveadditionalfundinginjectionscouldsignificantlyaccelerateadoptionbeyondtheseprojections.

Tofurtherspeedupthedeploymentofvehicle

autonomy,theindustryneedstokeepworking

onfivedifferentfronts.First,bringthepublicon

boardbycommunicatingconsistentmessages

andbuildingconsumertrust.Second,continue

leveragingadvancesintechnology,includingAI

andcybersecuritybreakthroughs,totacklethe

currentshortcomingsurroundingsafety,usabilityandscalability.Third,developsustainablebusinessmodelsthatfosterlong-termviability.Fourth,

co-createregulationstohelppolicymakersbetterunderstandtheprogressandreadinessofvehicleautomationtechnology.And,finally,collaboratewithinandacrossindustriestobetterfacilitate

large-scaledeployments.

AutonomousVehicles:TimelineandRoadmapAhead4

Introduction

Autonomousvehiclesmovebeyond

initialhypeanddisillusionmenttowardsreal-worlddeployment.

Thiswhitepaperaimstoshedlightontheevolving

vehicleautonomytimelinebetweennowand2035.

Theforecastsdevelopedherearebasedonanalysisgroundedinfivekeydimensions:

1.Consumertrustandinterest

2.ProjectedADAS/ADpricesandconsumers’willingnesstopay

3.Technologicalobstaclesandthetimeframeforovercomingthem

4.Currentregulatorystatusandanticipatedregulatorydevelopments

5.Ecosystemdevelopmentstosupportscaling

Whilethefirsttwodimensionshelpdetermine

thepotentialdemand,thelastthreeconcernthepotentialsupply.Thesedimensionsalsoformthebasisoftheactionsoutlinedinthispapertosafelyscalevehicleautonomy.

Theforecastsinthiswhitepaperaimtoaccountforexpectedprogress.However,technological

breakthroughs,suchasthesuccessful

deploymentofmap-freeandvision-onlyL3/L4

systems,ormassiveadditionalfundinginjectionscouldsignificantlyaccelerateadoptionbeyondtheseprojections.

Untilveryrecently,forecastsstatedthat

autonomousvehicleswouldbeeverywhere

inthe2020s.However,itisnowevidentthat

technological,regulatoryandeconomicchallengesmeanadoptionwillhappenmoregradually.

Despitethedifficulties,therationaleforAVsremainsstrong,drivenbypotentialsafetyandefficiency

benefits,amongothers.Roadsafetyremainsoneofthemostpressingconcernsinglobaltransport,andadvanceddriverassistanceandautonomousdriving(ADAS/AD)couldhelpreducethe1.2millionroadfatalitiesthatoccureachyear.1Themany

efficiencygainsincludeachievinghighervehicle

utilizationratesthroughreducedidletimeand

maximizedvehicleloads.Theseadvantagesapplytobothpassengertransport,throughridepooling,andgoodstransport,throughoptimizedfreight

movement.Bettertransportoptionsmayalsospurashiftawayfrompersonalvehicledependence,

leadingtoamoresustainablemobilitysystem.

Thiswhitepaperaimstoshedlightontheevolvingvehicleautonomytimelinebetweennowand2035acrossthreekeyusecases:personalvehicles,

robotaxisandautonomoustrucks(seeFigure1).

Special-purposevehiclesoperatinginenclosed

facilities,suchasthoseforminingoragriculture,arealsohighlysuitedforautomationbutfalloutsideofthescopeofthiswhitepaper.

FIGURE1Comparativeoverviewofthefourmainvehicleautonomysegments

Specialpurpose

autonomousvehicles

–Improvesafetyinhazardousenvironments

–Enhanceef?ciencyforspecializedtasks

Specialist?rmsownandoperate

Focusofthiswhitepaper

Personalvehicles

Autonomoustrucks

–Addresscriticaldrivershortages

–Increaseef?ciencyand

?exibilitywith24/7uptime

Fleetprovidersownandoperate

–Increaseroadsafetybyreducinghumanerror

–Enhanceconvenienceduringtravel

Privatelyownedorleased

Robotaxisandroboshuttles

Expectedbene?ts

–Enhancethe?exibilityofpublictransport

–Reduceoperationalcostsandimproveaccessibility

Ownership

Fleetprovidersownandoperate

Techlevel*Gradualdevelopmentfrom

ADAS(L0-L2+)toAD(L3/L4)

Autonomy-?rstsystemdevelopment(L4)

Autonomy-?rstsystemdevelopment(L4)

Autonomy-?rstsystemdevelopment(L4)

DomainHighway,suburbanandurbanSuburbanandurbanHighwayandsuburbanSpecialenvironments

*SeeBox1formoreinformationontechnologylevels.

AutonomousVehicles:TimelineandRoadmapAhead5

BOX1DefinitionofADAS/ADlevels

wheelintheirhands,keeptheireyesontheroadorfocus

theirmindonthedrivingenvironment(e.g.,whethersleepingisallowed).

Level4isthefirstlevelconsideredautonomous.Thatis,

wherethedriverhasnodrivingtasksintheoperatingdesigndomainspecified.ForL4inprivatevehicles,thiswhitepaperdifferentiatesbetweenL4HighwayandL4Urban,since,duetothegreatercomplexitiesofurbanareas,ADcapabilitiesareexpectedtobecomecommonplacesoonerinhighwaysettings.

Vehicleautomationisclassifiedintodifferentlevels

thatdifferentiatetheextentofautomation,thedriver’s

involvementandwheretherespectivesystemcanbeused–theso-calledoperationaldesigndomain(ODD).This

classificationreflectsthetechnologicalprogressandhelps

ensureclarityaboutthetechnology’scapabilitiesandliability.

Tomakethetechnologicaldifferencesevenclearer,the

levelsarefurthersubdividedaccordingtothedriver’srequiredactivities,namelywhetheritisrequiredtoholdthesteering

Assisted

Automated

Autonomous

Samplefeatures

–Automaticemergencybraking

–Lanedeparturewarning

–Adaptivecruisecontrol(ACC)

–Lane-keepingassistsystem(LKAS)

Explanation

–Safetywarningsor

temporaryassistance

–Driverretainsalldrivingtasks

L0

Manual

–SteeringORspeedcontrolbythesystem

–Driverremainshands-onandeyes-on

L1

Assisteddriving

–SteeringANDspeedcontrolbythesystem

–Driverremainshands-onandeyes-on

–CoupledACC&LKAS

L2

Partially

automateddriving

–SteeringANDspeedcontrolbythesystem

–Driverremainseyes-on

–Systemdrivesunderpre-definedconditions

–Driverneedstostepin

within~10secondsuponsystemrequest

–Systemdrivesunderpre-definedconditions

–Notake-overbythedriverisrequired(withintheODD)

–Systemdrivesinallconditions

–Notake-overbythedriverisrequired

L2+/++

–Navigateon

Autopilot(NOA)

Drivermustbeableto

immediatelytakefull

controlwheneverrequested

–Trafficjampilot

–Valetparking

Criticalchange:liabilityswitchesfromthedrivertothesystem

–AutonomousdrivinginapprovedODDs

Candifferentiate

betweenL4HighwayandL4Urban,duetotheir

differentcomplexities

–Ubiquitousautonomousdriving

Advancedpartially

automateddriving

L3

Automated

drivingunderconditions

L4

Autonomousdrivingunderconditions

L5

Autonomousdrivinginallconditions

SixlevelsofADAS/ADsystems

Hands-on

Eyes-on

Mind-on

System

.Driver

Source:Authors,adaptedfromthesixlevelsofvehicleautomationdefinedinSAEJ3016.2

AutonomousVehicles:TimelineandRoadmapAhead6

4%

Only4%ofnew

personalvehiclessoldin2035areexpectedtofeatureL4capabilities.

Assisted,automatedandautonomous

personalvehicles

PrivateADAS/ADadoptionisan

evolution,withpartiallyautomatedvehicles,notautonomousvehicles,dominatingthenextdecade.

PersonalvehiclesrepresentthelargestAV

marketsegmentbyvolume.Thismeansthat

automationadvancesinthisareaarecrucial

forimprovingroadsafetyandenhancingtravel

convenience.However,vehicleautomationin

personalvehiclesisanincrementalevolution

ratherthanadisruptiverevolution.Whilemany

projectionshavelongstateddriverlessvehicles

areimminent,thiswhitepaperfindsthat,overthenextdecade,personalvehicleswillbenefitprimarilyfromadvancedassistancefeaturesratherthan

autonomy.Detailedexpectationsfortheuptake

ofeachlevelbetween2023and2035areshown

inFigure2.Aswellaspersonalcars,Figure2alsoshowstheshareofL4robotaxis(moreonrobotaxisinthenextchapter).

Forecastsfortheuptakeofeachautomationlevel

Assisteddrivingtechnologies,particularlyatL2

andL2+levels,areexpectedtobethemost

dominanttechnologiesinnewcarssoldin2030

andbeyond.Thisisduetobroadmarketavailability,lowregulatoryhurdlesandlowersystemcosts

thanforL3/L4.Asaresult,driversofmostnew

carswillstillberequiredtokeeptheirhandson

thesteeringwheelandtheireyesontheroadlongafter2035.ThetransitionfromL2toL2+willalsobegradual,withL2systemsstayingmorepopularduringtheupcomingdecade.Thisisprimarilyduetocostconstraints:whileL2systemsarecurrentlyavailableforlessthan$700,L2+systemscancostupto$3,000.

L3vehicles–whereliabilitymovesfromthedrivertothesystem,yetthedriverstillneedstobeabletoregaincontrolinashorttimeframeifrequired–willremainatransitionalofferingratherthana

long-termsolution(asidefromspecificusecasessuchasvaletparkingortrafficjams).Therearefourkeyconstraintslimitingthewidespreadadoption

ofL3vehiclesthatarealsobeyondthetimelineconsideredinthiswhitepaper:

1.Safetyrisks:atL3,driversarerequired

toretakecontrolwithinaround10seconds,creatingpotentialsafetyconcernsinreal-worldconditions.

2.Liabilityconcerns:theshiftfromdriverresponsibilitytoOEMliability,withthe

relatedissueofdeterminingwhowasdriving,makesmanufacturershesitanttoscaleL3

inpublicenvironments.

3.Highcosts:L3requiresasimilartechstacktoL4,leadingtonearlyidenticalsystemcoststhattypicallyrangefrom$7,000to$10,000.

4.Limitedvalue:consumersmightexpectfullautonomy,buttheyarenotallowedtofullydisengageastheymayberequiredtotakeoverquickly.

Forthesereasons,manyOEMscontinueto

prioritizeL2/L2+advancementsoverL3,meaningthatby2035andpotentiallybeyond,L3vehicleswillstillconstituteonlyasmallfractionofsales.

Forpersonalvehicles,L4autonomyremains

significantlyconstrainedbybothtechnicalandregulatorybarriers.AtL4,ifthevehicleoperateswithinitsdefinedODD,thedriverisentirely

uninvolved,andthevehiclecanbringitselftoa

safestopifrequired.Only4%ofvehiclessold

in2035areexpectedtofeatureL4capabilities,

reflectingtheslowandselectivedeploymentoftrueautonomoustechnologyinthisarena.Inaddition,giventheadditionaloperationalchallengesinurbansettings,someofthesecapabilitieswillonlybe

availableforhighwayenvironments.Furthermore,inurbansettings,fleet-basedmodels,such

asrobotaxiservices,mayprevailoverprivateownershipinthelongerterm.

Finally,therearestillexpectedtobeasignificant

numberofnewvehicleswithL0andL1technologiesby2035.Thefeaturesofthesetechnologiesare

alsobecomingincreasinglycommoditized,withmanycountriesmandatingbasicsafetyfeatures,

1

AutonomousVehicles:TimelineandRoadmapAhead7

breakthroughsinAI(seeBox2),sensorsand

computingpower,offersthestrongestpotential

foracceleratingtherolloutofAVs.Yetevenifthistechnologyleapwasfullyrealized,L4technologieswouldstillonlybepresentinamodest7.5%of

newcarsalesby2035.Intermsofthelimitsto

uptake,theriskofunfavourableregulationislikelytobethebiggestbarrier.Suchregulationscouldbeenactedifmajor,negativeincidentsinvolvingL4technologiesoccurred,therebyunderliningtheimportanceofindustrycollaborationtoensurethesafetyofthesetechnologies.

suchaslane-keepingassistanceandautomatic

emergencybraking.L1systems,however,struggletodeliveracompellingvalueproposition:theyareasimilarcosttoL2systemsbutoffersignificantlylessfunctionality.Assuch,muchoftheL0marketshareisprojectedtoshiftdirectlytoL2overtime,bypassingL1systemsaltogether.

SupplementingthedatarepresentedinFigure2,

Figure3summarizesthesensitivityanalysiscarriedoutacrossthefivekeydimensionsunderlyingtheforecast.Technologicalprogress,drivenbypossible

Data,togetherwithbreakthroughsinE2EAIandscalablesoftware,areintegralfactorstofast-trackvehicleautonomyandenableglobaldeployment.

BerndSchmaul,ChiefDigitalOfficer,BoschMobility

FIGURE2PassengercarADAS/ADforecast

Globalnewcarsalesbyvehicleautonomylevel

<1%

13%

21%

39%

42%

13%

38%

12%

10%

44%

33%

24%

2023202520302035

L4HighwayandUrban

Note:Someofthetotalsdonotsumto100%duetorounding.

<1%

1%

1%

5%

1%

3%

4%

33%

13%

53%

L0

.L3

L1L2

●Robotaxis

L2+

L4Highway

FIGURE3Sensitivityanalysisofthepassengercarforecast

SupplyDemand

AutonomousVehicles:TimelineandRoadmapAhead8

PercentagepointchangeofL4

newcarsalesby2035fromthe

Dimension

Easy-to-usedeploymentswithpositivemediacoverageincreasedemandHigh-pro?leaccidentsnegativelyimpactpublictrust

Intensecompetition(incl.pricewars)drivesavailabilityofcost-ef?cientsolutionsPricesremainhighduetocomponentshortagesandnon-scalablesolutions

Technologicalbreakthroughsenableearlier,morescalableL4deploymentsKeyODDsremainunresolved,andscalabilitycontinuestobelimited

GovernmentssupportAVsfortechnologicalleadership,safetyandef?ciencyRegulatorslimitdeploymentduetomajorincidents

AVsolutionsscalegloballyacrossautomakers,standardizingtechPoorcoordinationandweakbusinesscaseshinderprogress

-1.0

+0.5

+3.5

-3.0

+1.5

-3.5

Consumertrustandinterest

Pricingand

willingnesstopay

Technologicalmaturity

Regulatory

developments

Ecosystemreadiness

Majorfactorsthatcouldin?uenceL4carsalesby2035

basescenarioof4%(seeFigure2)

+2.0

+1.0

-2.0

-2.0

BOX2

WhatdoadvancesinAImeanforvehicleautomation?

AIandgenerativeAI(GenAI)arebecomingintegraltovehicleautomationtechnology.Theyare

transformingdecision-making,modeltrainingandhuman-machinecollaboration,particularlyacrossthreekeyareas:3

1.End-to-end(E2E)AImodelsarereplacing

traditionalrule-basedsystemsthatstruggle

tohandlereal-worlddrivingcomplexity.

Bycombiningperception,predictionand

planningintoasingleneuralnetwork,E2EAIenablesfasterlearningandbetterresponsesacrossavarietyofenvironments.While

historicallycriticizedforalackoftransparency,recentinnovationismakingtheseE2EAI

modelsmoreinterpretableandverifiable,resolvingsafetyconcernsandincreasingindustryadoption.

2.Bycreatingsyntheticdata,GenAIplaysakeyroleintrainingautonomoussystems.Real-worlddatacollectioniscostlyand

inexhaustive,whereassimulationsusing

GenAIcreatemorediverse,scalabledatasetsthatcanexposemodelstounusualdriving

scenarios.Thisallowsautonomoussystems

tolearnfrommillionsofsimulatedmiles,

improvingtheirabilitytohandleedgecases,

suchassuddenobstaclesorextremeweather.

However,real-worldvalidationremains

essentialtoensurerobustnessandsafety.

3.AIisstrengtheninghuman-machine

collaborationthroughenhanceddriver

monitoringsystems(DMS)andhuman-

machineinterfaces(HMI).DMSuseAItotrackdriverattention,fatigueandstress,triggeringalertsorinterveningtopreventaccidents.

GenAIhelpsimprovevehicleinterfaces,

enablingmoreintuitivevoicecommandsandadaptivecontrolsthatminimizedistractions.

Moreover,byleveragingGenAI,systems

becomemorecapableofexplainingtheir

decisions,improvingtheuser’sunderstanding.

Regionaladoptiondifferences

AsFigure4demonstrates,personalvehicle

automationwillbeadoptedatdifferentratesaroundtheworld.TheshiftisexpectedtobeledbyChinafollowedbytheUnitedStates,withEuropeand

Japanlikelytofollowasimilaryetslowertrajectoryoverthecomingdecade.IntermsofL2+adoption,theshareofnewcarsalesin2035isexpected

tobesignificantlygreaterinChinathananyothermajorterritory.Thisiscaused,inpart,bythe

willingnessofChinesecustomerstoembrace

automationanddomesticOEMsandsuppliers’rapidadvancesinautomation.Chinaisalso

expectedtohaveslightlyhighersharesofL3andL4vehiclesthanothergeographiesin2035–the

USandEuropebeingtheonlyothertwomarketswherethesetechnologieswillstarttoappearinprivatelyownedpassengercarsby2035.

Beyondthesefourkeymarkets,therestof

theworldfollowsamixedtrajectory,withsome

regionssteadilyadoptingL2andL2+systems

whileothersfaceeconomic,technologicaland

regulatoryhurdlesthatwillslowthetransitionto

higherADAS/ADlevels.Figure4alsohighlights

theexampleofIndia,whichfollowsasomewhat

differenttrajectorytotheotherhighlightedmarkets.

InIndia,L0systemsareforecastedtostilldominatein2035duetolowerpurchasingpowerandmorecomplexroadenvironments.IndiaisalsoexpectedtoleapfrogL1andmovemoredirectlytoL2as

thistechnologymatures.

Autonomousdrivingtransformscarsintolivingspaces.ADAS/ADsystemsmustbeaccessibleacrossallregionsandsegments,enhancingsafetyanduser-centricin-vehicleexperiencesforeveryone.

GürcanKarakas,CEO,TOGG

AutonomousVehicles:TimelineandRoadmapAhead9

FIGURE4PassengercarADAS/ADforecastbyregion

Newcarsalesbyvehicleautonomylevel

China

Europe

1%

1%

6%

8%

8%

9%

23%

10%

39%

25%

46%

41%

39%

53%

7%

37%

27%

6%

11%

0%

1%

7%

19%

23%

22%

10%

18%

24%

32%

49%

49%

53%

45%

15%

28%

UnitedStates

2%

6%

13%

4%

46%

48%

45%

5%

21%

17%

17%

17%

40%

35%

29%

24%

14%

14%

202320252030203520232025203020352023202520302035

India

Japan

2%

5%

8%

50%

15%

29%

12%

59%

1%

1%

11%

45%

11%

14%

51%

15%

35%

43%

41%

50%

12%

13%

12%

13%

44%

41%

36%

32%

52%

28%

19%

RestofWorld

5%

6%

1%

13% 3%

23%

4%

95%

93%

84%

73%

2023202520302035202320252030203520232025