Automated Driving Systems (ADSs) have the potential to make mobility services both accessible and safe. A multi-pillar Safety Assurance Framework (SAF) has been proposed to support the safety assessment of ADSs, requiring comprehensive test scenario coverage of the system’s Operational Design Domain (ODD). This work addresses two key questions: first, how a Key Performance Indicator (KPI) can be evaluated within the proposed SAF, and second, how heterogeneous test environments contribute to a comprehensive KPI assessment. By focusing on specific KPIs, this paper can demonstrate the complete walk-through of how external requirements for the system under test (SUT) are integrated into the SAF, leading to the generation of concrete test scenarios based on the ODD, dynamic driving task (DDT), and test objectives. The allocation of test cases to relevant test environments, both virtual and physical, follows the proposed methods and the evaluation of safety criteria. The use case is automated truck parking with a semi-trailer in confined logistics environments, such as terminals, providing a controlled setting to validate highly automated vehicles’ low-speed perception and decision-making capabilities (L4). This use case employs simulations and physical tests, using model-scale and full-size trucks, to showcase the SAF’s ability to guide a safety assessment.
Skoglund et al. (Tue,) studied this question.