WIXLIB

the latest status in Austria and Spain is presented. Further the infrastructure data can be alignedwith automotive safety integrity level.The users’ driving and travel experience will be explored to provide an incentive for OEMs tocooperate with the infrastructure, since it can provide additional information for on-boarddecisions of CAVs. This requires answering the question as what the prerequisites towardsthe infrastructure from vehicle side are. The most basic step, a classification of infrastructureis needed, which ISAD levels can provide. Still, a common understanding between OEMs,automotive industry and road operators is to be established.2.3.1. ISADThe environmental perception of automated vehicles is limited by the range and capability of on‐board sensors. Road infrastructure operators already employ numerous traffic and environmentalsensors and provide information that can be perceived by automated vehicles. In order to classifyand harmonize the capabilities of a road infrastructure to support and guide automated vehicles,we support a simple classification scheme, similar to SAE levels for the automated vehiclecapabilities, following the work of the EU research project INFRAMIX2. This project is preparingthe road infrastructure to support the transition period and the coexistence of conventional andautomated vehicles. Its main target is to design, upgrade, adapt and test both physical and digitalelements of the road infrastructure, ensuring an uninterrupted, predictable, safe and efficienttraffic. Although INFRAMIX is targeting mainly highways, since they are expected to be the initialhosts of such mixed traffic, its key results can be easily transferred to urban roads.The levels we support and which are being developed in INFRAMIX are called InfrastructureSupport levels for Automated Driving (ISAD) and can be assigned to parts of the network in orderto give automated vehicles and their operators guidance on the “readiness” of the road networkfor the coming highway automation era.ConventionalinfrastructureCDEBased on the real-time information on vehicle movements, theinfrastructure is able to guide AVs (groups of vehicles or singlevehicles) in order to optimize the overall traffic flow.Infrastructure is capable of perceiving microscopic traffic situationsCooperative perceptionand providing this data to AVs in real-timeAll dynamic and static infrastructure information is available inDynamic digital informationdigital form and can be provided to AVs.Digital map data is available with static road signs. Map data couldStatic digital information / be complemented by physical reference points (landmarks signs).Map supportTraffic lights, short term road works and VMS need to be recognizedby AVs.Conventional infrastructure / Conventional infrastructure without digital information. AVs need tono AV supportrecognise road geometry and road signs.Cooperative drivingGuidance: speed,gap, lane adviceBDescriptionMicroscopic trafficsituationANameVMS, warnings,incidents, weatherDigitalinfrastructureLevelDigital map withstatic road signsFigure 2 “ISAD levels” shows the description of the five ISAD levels, along with the informationthat is provided to automated vehicles.XXXXXXXXXXFigure2 "ISADLevels"Figure2: "ISADLevels"2Infrastructure support levels are meant to describe road or motorway sections rather than whole roadnetworks. This reflects common practice of infrastructure deployment: Traffic control systems (sensorsand VMS) are usually deployed on motorway sections where traffic often reaches the capacity limit(e.g.in metropolitan areas), whereas other motorway sections need no fixed installations of traffic https://www.inframix.eu/control systems because traffic flow is rarely disrupted. The five levels increase in added support from7

ConventionalinfrastructureABCDEBased on the real-time information on vehicle movements, theinfrastructure is able to guide AVs (groups of vehicles or singlevehicles) in order to optimize the overall traffic flow.Infrastructure is capable of perceiving microscopic traffic situationsCooperative perceptionand providing this data to AVs in real-timeAll dynamic and static infrastructure information is available inDynamic digital informationdigital form and can be provided to AVs.Digital map data is available with static road signs. Map data couldStatic digital information / be complemented by physical reference points (landmarks signs).Map supportTraffic lights, short term road works and VMS need to be recognizedby AVs.Conventional infrastructure / Conventional infrastructure without digital information. AVs need tono AV supportrecognise road geometry and road signs.Cooperative drivingGuidance: speedgap, lane adviceDigitalinfrastructureConnected Automated Driving RoadmapMicroscopic traffisituationDescriptionVMS, warnings,incidents, weatheNameDigital map withstatic road signsLevelXXXXXXXXXXFigure 2 "ISAD Levels"Infrastructuresupport levels are meant to describe road or motorway sections rather than wholeroadInfrastructurenetworks. Thisreflectsinfrastructurecontrolsupportlevels commonare meant practiceto describeofroador motorwaydeployment:sections ratherTrafficthan orwaysectionswheretrafficoftenroad networks. This reflects common practice of infrastructure deployment: Traffic controlreachesthe (sensorscapacityandlimit(e.g. arein usually deployedon whereasmotorwayothersectionswhere trafficoftenno fixedinstallationsoflimittrafficcontrolsystems becausetraffic flowrarely disrupted.reachesthe capacity(e.g.in metropolitanareas), whereasotherismotorwaysections Theneedfivelevelsin added ofsupportfrom E systemsto A. Thisexample(Figure3) rarelyillustrateshow ISADlevelsno increasefixed installationstraffic controlbecausetrafficflow isdisrupted.The fivecan levelsbe usedfor a simpledescriptionof whatvehiclescan expecton specificparts ofincreasein addedsupport fromE to automatedA. This example(Figure3) illustrateshow ISADlevelscan be used for a simple description of what automated vehicles can expect on specifica roadnetwork.parts of a road network.Figure 3 “ISAD Example”Figure3: “ISAD Example”2.4 Regulatoryandstandardisation framework for AutomationThe vehicle operates under two major regulations: the homologation framework and the onframeworkfor toAutomationframework.new concepts of ODDand ISAD needcomply with theconstraints defined by these frameworks. The status today forms constraints for theThe vehicle operates under two major regulations: the homologation framework and the trafficregulation framework. The new concepts of ODD and ISAD need to comply with the constraintsdefined by these frameworks. The status today forms constraints for the implementation ofautomation.It- isclearthat both frameworks at the same time have to evolve themselves to caterwww.ertrac.orgTwitter- LinkedInPage 9 of 56properly for future traffic with higher automation levels and mixed traffic.Additional research needs to be carried out and a joint approach between telecom and vehicleindustry as well as cross-border pilot projects and the adaptation of road traffic rules in MemberStates can all support to reach a consensus in this field. The deployment of CAD may behampered if there are no collaborative actions towards a coordinated and quick development of8

implementation of automation. It is clear that both frameworks at the same time have to evolvethemselves to cater properly for future traffic with higher automation levels and mixed traffic.Additional research needs to be carried out and a joint approach between telecom and vehicleindustry as well as cross-border pilot projects and the adaptation of road traffic rules in MemberEuropean-wide regulations and laws enabling testing and use of automated vehicles on publicStates can all support to reach a consensus in this field. The deployment of CAD may beroads. By solving regulatory issues, the public will start to accept and use automated vehicleshampered if there are no collaborative actions towards a coordinated and quick development ofsooner, leading to a better market penetration and a competitive advantage for the EU on theregulations and laws enabling testing and use of automated vehicles on publicfield ofEuropean-wideautomated driving.roads. By solving regulatory issues, the public will start to accept and use automated vehiclesleadingto a thebettermarketpenetrationandwhicha competitivetheonEU on aintsundera vehicle isadvantageallowed toformovefieldcoveringof automatedthe road,aspectsdriving.like speed, allowed vehicle characteristics like width, height, constraintswhichetc.a vehicleis allowed to moveand permissible movements like laneright/leftturn, underovertaking,Traffic regulationson the road, ofcoveringaspects level,like speed,vehiclecharacteristicslikeapply independentthe automationso theyallowedcan providean importantinput tothewidth,ODD height,assessment.weight and permissible movements like lane change, right/left turn, overtaking, etc. Trafficregulations apply independent of the automation level, so they can provide an important inputTraffictoregulationsare implemented traditionally in the ISAD level E world via “road signs”the ODD assessment.placedTrafficon, aboveor near thewhere thetraditionallyterm sign tedin theroadISADlevel E Vehiclesworld via“road signs”traffic ectionprobabilitymaybelimitedinon, above or near the road, where the term sign includes road markings. thetraffic regulations via sensors (e.g. cameras), but the detection probability may be limited incurrently valid traffic regulations at the same time form an important part of the ISAD content.difficult environments (adverse weather conditions, sign “forest” in urban scenarios), hence theIn future Level 5 scenarios independent of traditional visualisation means more dynamic trafficcurrently valid traffic regulations at the same time form an important part of the ISAD content.regulations become possible – e.g. access regulations based on certain vehicle criteria that onlyIn future Level 5 scenarios independent of traditional visualisation means more dynamic trafficapply if certain conditions are met – which would make traffic regulations as an element of ISADregulations become possible – e.g. access regulations based on certain vehicle criteria that onlymandatory.apply if certain conditions are met – which would make traffic regulations as an element ofISAD mandatory.Figure 4 Regulatory and standards framework for AutomationElaborationof ODDand ISADshouldforbeAutomationperformedFigure4: Regulatoryand standardsframeworkcollectively in a pre-competitiveenvironment between automotive and infrastructure sectors. ISAD will provide importantelementsfor andODDdefinitions,regarding collectivelydigital infrastructure(e.g. availabilityof electronicElaborationof ODDISADshould be performedin a pre-competitiveenvironmenttrafficregulation)as well as physicalinfrastructure(e.g.about elementsroad markingquality orbetweenautomotiveand infrastructuresectors. ISADwill provideimportantfor ODDdefinitions,regardingdigitalinfrastructure(e.g. Theavailabilityof electronicas based onstandardisedroadlayoutsin roadworks).ISAD interfaceneedstrafficto be regulation)standardisedwell asODDphysicalinfrastructure (e.g. about road marking quality or standardised road layouts inrequirements.roadworks). The ISAD interface needs to be standardised based on ODD requirements.9www.ertrac.org - Twitter - LinkedIn

Connected Automated Driving RoadmapConnectivity asasa requirementfor tya requirementfor vehicle-infrastructureinteractionWhilstthe automation side is heavily discussed, one should not overlook the connectivityaspectsof CAD. Forsidehigherautomationlevels, oneconnectivityis aoverlookpivotal aspectgiven theWhilstthe automationis heavilydiscussed,should notthe dparties.Toensureaproductivedevelopmentofaspects of CAD. For higher automation levels, connectivity is a pivotal aspect given theCAD, each participant must also be aware where he is located not only geographically but alsonecessaryinteractions between all involved parties. To ensure a productive development ofdigitally in relation to the other stakeholders in this connected network. The understanding andCAD,each participant must also be aware where he is located not only geographically but alsodevelopment of connectivity can and should of course be further elaborated and supported bydigitally in relation to the other stakeholders in this connected network. The understanding andthe involved stakeholders.development of connectivity can and should of course be further elaborated and supported bythe Theinvolvedstakeholders.currently developed communication profiles and services connect all involved parties viaTheshortcurrentlycommunicationprofiles thisandconnectivityservices connectall them,involvedpartiesrangedevelopedcommunication.To further improvebetweenthereis rovethis connectivitybetweenthem,differentthere is alsotherangeoptionto support hybridsame contentcan be sentout throughthe media,option compatibleto support fferentmedia,and interoperable.compatible and interoperable.An Anoverviewdata rviewofof thethe datanationalroadroadauthoritiescan becanseenFigurein5 Figure .“Stakeholdersin inNRA5: “Stakeholdersin NRAdata exchange”Figure 5Figure“Stakeholdersin NRA dataexchange”10This can further be supported by multi-channel communication to upgrade existing ITS servicesusing conventional standard system architectures and message formats. It is important here touse standardised mappings from C-ITS data structures to existing standards (e.g. DATEX II) toensure interoperability. This adds another layer to connectivity.A major factor for the future evolution of connectivity will be the emergence of 5G networks.The 5G technology will be much more powerful than today’s 3G/4G networks. In the currentpractice mobile networks are mainly used as simple access networks to the mobile internet.Mobile stations (vehicles, smartphones, etc.) have point to point links to services. Althoughproposed hybrid communication architectures de-facto provide geographical addressing viamessage brokers, the underlying communication technology is still based on point to pointcommunication. 5G may potentially change this with sophisticated architectural features likemobile edge computing and unprecedented bandwidth and latency characteristics. This mayallow connectivity patterns that go beyond what is possible today.

This can further be supported by multi-channel communication to upgrade existing ITS servicesusing conventional standard system architectures and message formats. It is important here touse standardised mappings from C-ITS data structures to existing standards (e.g. DATEX II) toensure interoperability. This adds another layer to connectivity.A major factor for the future evolution of connectivity will be the emergence of 5G networks.The 5G technology will be much more powerful than today’s 3G/4G networks. In the currentpractice mobile networks are mainly used as simple access networks to the mobile internet.Mobile stations (vehicles, smartphones, etc.) have point to point links to services. Althoughproposed hybrid communication architectures de-facto provide geographical addressing viamessage brokers, the underlying communication technology is still based on point to pointcommunication. 5G may potentially change this with sophisticated architectural features likemobile edge computing and unprecedented bandwidth and latency characteristics. This mayallow connectivity patterns that go beyond what is possible today.11

Connected Automated Driving Roadmap3. DEVELOPMENT PATHSStakeholders involved in ERTRAC share the vision of a progressive step-wise increase ofautomation levels during the upcoming decade. The main development paths for the differentautomation levels are shown in the Figure 2 below.Major collaborative projects funded by the EU are currently looking at the first levels ofautomation and organizing testing in order to prepare their roll-out in Europe: the L3Pilot3 projectfor automated passenger cars, and the ENSEMBLE4 project for multi-brand truck platooning.These projects were strongly promoted by the previous versions of the ERTRAC roadmap. Thefocus for this version of the roadmap is on High Automation - Level 4.The development paths indicate the timeframe to reach levels of TRL 7-9, to be understood as ready froma technology perspective. We refer to the TRL levels definition used within the Horizon 2020 programme5: TRL 7 – system prototype demonstration in operational environment TRL 8 – system complete and qualified TRL 9 – actual system proven in operational environmentAutomation LevelEstablished20182020202220242026Level 5:Full Automation.Auto PilotLevel 3:ConditionalAutomationChauffeurLevel 2:Partial AutomationLevel 0:No DrivingAutomation, supportbeyond humancapability to act2030Full AutomationLevel 4:High AutomationLevel 1:Driver Assistance2028Driving AssistanceAdvanced DriverAssistance Systems(ADAS)Warning or Supportby active safetysystemsFigure 6: The vehicle automation development paths31245 https://www.l3pilot.eu https://platooningensemble.eu f/h2020/wp/2014 2015/annexes/h2020-wp1415-annex-g-trl en.pdf

Since road transport includes various types of vehicles, it is important to detail the developmentpaths into specific roadmaps reflecting the different opportunities of each vehicle category.The following subchapters present three different roadmaps providing a specific outlook for:passenger cars, freight-vehicles, and urban mobility vehicles and buses.3.1. Automated Passenger Cars PathPassenger cars are the main driver of the development towards automated driving, as with theirhigh volume in the market, they can afford to develop the necessary technologies. They evolvelevel by level with more sensors, connectivity and computing power on- and off-board and canbe distinguished by parking and driving use cases. All level 0 to 2 systems definitions have nowbeen placed in the Annex of this roadmap version, as well as the automated parking assistancesystems.Automated Passenger Car Development PathsAutomation LevelEstablished2018202020222024202620282030.Level 5:Full AutomationFully AutomatedPassenger CarsUrban and Sub-Urban PilotLevel 4:High AutomationLevel 3:ConditionalAutomationHighway Autopilot includingHighway ConvoyHighway ChauffeurTraffic Jam ChauffeurLevel 2:Partial AutomationTraffic Jam AssistParking AssistLevel 1:Driver AssistanceAdaptive Cruise ControlStop &GoLane Keeping AssistLane Change AssistParking assistLevel 0:No DrivingAutomation, supportbeyond humancapability to actLane Departure WarningBlind-spot WarningForward Collission WarningABS, ESCEmergency BrakeIn dic ativeede vlopmenthst paTR L7- 9Passenger Cars: M1 categoryFigure 7: The Automated Driving development path for passenger cars13

Connected Automated Driving Roadmap3.1.1. Traffic Jam Chauffeur (Level 3)Conditional automated driving in traffic jam up to 60 km/h on motorways and motorway similarroads. The system can be activated in case of a traffic jam scenario. It detects slow drivingvehicle in front and then handles the vehicle’s both longitudinal and lateral. Later version of thisfunctionality might include lane change functionality. The driver must deliberately activate thesystem, but does not have to monitor the system con

2.1. Levels of Automation 5 2.2. Operational Design Domain 6 2.3. Vehicle and infrastructure interaction 6 2.4. Regulatory and standardisation framework for Automation 8 2.5. Connectivity as a requirement for vehicle-infrastructure interaction 10 3. Development paths 12 3.1. Automated Passenger Cars Path 13 3.2. Automated Freight Vehicles Path .