VErification methods for Robust Driver Assist System performance (VERDAS)

Purpose and goal:

The goal of the project is to support significantly improved road safety for automated vehicles by developing physical and virtual verification methods for efficient development of Driver assistance systems to take care of a larger part of all accident scenarios and that the safety function performance should be robust in real-world accident scenarios. Even if retrospective studies show reduction of seriously injured and killed it still shows function limitations despite many years of development of Driver assistance systems and verification methods e.g. in Euro NCAP.

Expected results and effects:

Euro NCAP has identified those function limitations in Vision 2030. The first introduction of verification methods for robust Driver assistance systems will be introduced in Euro NCAP rating 2026. To enable the verification of a larger number of all accident scenarios, virtual testing is required to complement physical testing.

The rating will include a wider range of real-world accident test scenarios and perception layers with e.g. traffic environments and other road users. VERDAS will support Euro NCAP and have a specific focus on accident scenarios with vulnerable road users. Approach and implementation:

AstaZero AB will be leading the project and the partners are If Insurance, Folksam, Trafikverket, Volvo Car Corporation, Toyota, Zenseact, Aptiv, Terranet Tech AB and Viscando AB. The project will be conducted during May 2024 - May 2025 and will include different work packages with Field data analysis, Definition of Robustness test scenarios, Definition of Robustness test methods and development of Perception layer, Virtual verification methods, Verification tests at test track of robustness test methods and also Physical/Virtual test protocols, all in dialog with Euro NCAP.