Efficient safeguarding of autonomous driving functions with HolmeS³

As the complexity of autonomous driving functions increases, so do the challenges of validation. Traditional test methods are no longer sufficient to cover all possible scenarios in real road traffic. The HolmeS³ research project, a collaboration between the Ostbayerischen Technischen Hochschule, e:fs TechHub GmbH and imbus AG, offers an innovative solution here.

HolmeS³ combines causal models with scenario-based testing to specifically identify critical scenarios and efficiently generate test cases. Causal inference analyses cause-and-effect relationships and systematically derives relevant corner cases. This allows tests to be prioritised in a targeted manner and the test effort to be significantly reduced.

The solution for scenario-based testing developed in the HolmeS³ project (based on the imbus test management tool TestBench) enables the targeted selection of relevant corner cases for specific driving scenarios and their execution in the simulator. In this way, the behaviour under extreme conditions is checked and validated. The method fulfils the highest safety requirements in accordance with ISO 21448 (SOTIF) and improves the efficiency of testing highly automated driving functions.

Further details on the HolmeS³ project can be found here.

All about HolmeS³