Statistical Model Checking for Scenario-based verification of ADAS

Sebastian Gerwinn and Eike Möhlmann and Anja Sieper
Control Strategies for Advanced Driver Assistance Systems and Autonomous Driving Functions : Development, Testing and Verification
The increasing complexity of advanced driver assistant systems requires a thorough investigation of a high-dimensional input space. To alleviate this problem, simulation-based approaches have recently been proposed. An overall safety assessment, however, requires the simulation results to relate to the safety of such systems in real-world scenes. To this end, we propose a rigorous method of specifying requirements for these systems depending on the environmental situation and generating statistical evidence for the safety of the system in the specified environmental situations. We demonstrate this process in an exemplary highway scenario involving decision and perception uncertainty.
Springer International Publishing