Wednesday, June 28, 2023

Closed Loop Test Case Selection for the Safety Assessment of Automated Driving

Topic and Goal of the Thesis

Automated vehicles have a significantly higher operational scope than driver assistance systems. That is why research is focusing on a scenario-based approach for proving their safety. In this approach, traffic is broken down into individual scenario classes and the vehicle is specifically tested in these. In that way, it is also possible to specifically test particularly interesting scenarios that occur less frequently in real traffic.

An open research question is how to determine which scenarios found are suitable as test cases for an automated driving system (System Under Test, SUT). In this thesis, an approach shall be developed and implemented to select test cases efficiently. For this purpose, the SUT shall drive in a set of initial test cases in a simulation environment and based on its performance new suitable test cases shall be generated automatically.

Working Points

• Literature research on scenario-based testing
• Development of the closed loop selection method
• Implementation and evaluation in a simulation environment

Requirements

• Experience with Python
• Prior knowledge on automotive engineering, C++ are beneficial
• Reliability, commitment and enjoyment of working independently
• Good English or German language skills

Note: Please attach brief resume and grade summary.