Simulation of Abstract Scenarios: Towards Automated Tooling in Criticality Analysis

Abstract

While the introduction of automated vehicles to public roads promises various ecological, economical and societal benefits, reliable verification & validation processes that guarantee safe operation of automated vehicles are subject to ongoing research. As automated vehicles are safety-critical complex systems, operating in an open context, the uncountable infinite set of potentially critical situations renders traditional, distance-based approaches to verification & validation infeasible. Leveraging the power of abstraction, current scenario-based approaches aim at reducing this complexity by elic-itation of representative scenario classes while simultaneously shifting significant analysis and testing efforts to virtual environments. In this work we bridge the gap between high-level, abstract scenario specifications and state-of-the-art detailed vehicle and traffic simulators. While the first allow for coverage argumentation with the definition of finite and well manageable sets of scenario classes the latter is necessary for a in-depth assessment of the vehicle implementation and its interaction with the physical environment. We present a method and prototypical implementation based on constraint solving techniques to generate (sets of) concrete simulation tasks defined in the well established OpenDRIVE/OpenSCENARIO formats from abstract scenarios specified as Traffic Sequence Charts. The feasibility is demonstrated using a highway parallel overtaking scenario as a running example.