ExpectedOutcome:
Project results are expected to contribute to all of the following expected outcomes:
- Safe scaling up of the deployment of CCAM systems for all levels of automation, including systems that for part of the driving phases rely on human-machine interaction.
- Assurance of vehicle safety despite system changes, e.g., due to software updates and data exchanges between vehicles and the infrastructure.
- Facilitating the introduction of fast developing technological innovations in the CCAM system’s functionality, such as AI.
Scope:
To ensure the safety of CCAM, it is essential that vehicles are not only safe during the (first) type approval, but also during their complete lifetime in a fast-changing road transport system. Changes can result from the evolution of the CCAM system itself, for example, as a result of increasing connectivity using V2X communication, the use of AI-based systems, and OTA (over-the-air) software updates. The traffic system, in which CCAM systems are being deployed, is changing at a rapid pace as well, with an increased market share of vehicles with higher levels of automation, new (personal) mobility devices and autonomous mobility robots (e.g., for package delivery).
At the same time, the way CCAM systems interact with humans in traffic is changing. Until full automation in transport is reached, the human driver will keep on playing an essential role. Also, the interaction with other road users will change, supported by technologies that allow a CCAM system to communicate its intentions to other road users.
As a consequence of these innovations and developments, the safe deployment of CCAM systems needs an extension of the safety validation procedures and certification schemes, taking advanced human-machine interaction and a continuous in-service monitoring approach into account. Due to the many different scenarios and variations that can occur realistically and that consequently need to be tested, it should be possible that a large part of the assessment is performed in a virtual simulation environment.
The proposed actions are expected to address all of the following aspects:
- Developing a validation methodology for scenario-based safety assurance of AI-based CCAM functions. Trustworthiness of the AI-algorithms depends on how well the system responds to scenarios in its Operational Design Domain (ODD) – specificity and how it responds in case it ends-up outside its ODD – robustness. Consequently, methods need to be developed on the use of scenarios to describe the ODD of AI-based systems.
- Connectivity. Developing validation procedures for CCAM systems that rely on V2X for safety-critical functions i.e., the inclusion of the connectivity context. Ensuring aspects of reliability, trustworthiness and cyber-security with respect to V2X is essential. The approach to V2X connectivity is technology neutral.
- Continuous Safety Assurance approach. Developing an approach for a continuous safety validation methodology, to monitor the safety state of deployed CCAM systems in operation (real traffic) during its service life, following type approval. Performance metrics for the reliability of the monitored data, including cyber-security aspects, and indicators for the safety state should be proposed. Also needed is the development of requirements for the monitoring system for use in future standardisation, regarding the exchange of data and safety performance indicators with service organisations and authorities.
- Validating the virtual approach. Developing tools that ensure the relevant degree of detail and the appropriate representation of other road users’ behaviour (incl. Vulnerable Road Users such as pedestrians and/or bicyclists) in virtual scenario-based testing. This includes methods to deal with perception, localisation, and world modelling errors in the validation procedures.
- Human Machine Interaction. Developing a safety assurance methodology that incorporates the assessment of Human Machine Interaction (both driver-vehicle and vehicle-road user) concepts for higher levels of automation (conformity checks as well as test set-ups with suitable metrics) ensuring safe communication between driver and vehicle and between vehicle and other road users, making Human Machine Interaction inclusive (i.e. in terms of age, mental and physical ability, cultural aspects, etc.).
Proposed actions are expected to develop recommendations for harmonisation and standardisation and to feed into on-going discussions regarding EU type vehicle approval rules as well as in the framework of the UNECE.
Actions should be based on the outcomes of previous methodologies developed in HEADSTART[1], as well as research funded under HORIZON-CL5-2021-D6-01-02[2].
Upcoming CCAM projects, in particular in the area of large-scale demonstrations, validation, digital infrastructure and key enabling technologies should be taken into account to ensure compatibility.
Links should be established with the Mobility Data Space initiatives from Digital Europe, federated data infrastructure projects (Gaia-X, International Data Spaces, Big Data Value - BDV).
In order to achieve the expected outcomes, international cooperation is encouraged, in particular with Japan and the United States but also with other relevant strategic partners in third countries.
This topic implements the co-programmed European Partnership on ‘Connected, Cooperative and Automated Mobility’ (CCAM). As such, projects resulting from this topic will be expected to report on results to the European Partnership ‘Connected, Cooperative and Automated Mobility’ (CCAM) in support of the monitoring of its KPIs.
Specific Topic Conditions:
Activities are expected to achieve TRL 5 by the end of the project – see General Annex B.
[1]https://www.headstart-project.eu/
[2]“Common approaches for the safety validation of CCAM systems”
