Automotive Industry Insights

V&V Methods (VVM): Verification and Validation Methods for Automated Vehicles in Urban Environments

Project duration: 2019 – June 2023
Funded by the German Federal Ministry for Economic Affairs and Energy (BMWi).

SETLevel and VVM build on the results of the PEGASUS project, which was concerned with identifying and describing critical scenarios and turning them into universally applicable test cases for highly automated vehicles, using the Autobahn Pilot as an example. The Verification and Validation Methods project, or VVM for short, extends the PEGASUS method, taking a downtown junction as an example.

The approach mapped out for the VVM project can be described in three essential steps. Validating an autonomous vehicle by means of test drives in normal road traffic would require driving several million kilometers in order to cover enough different situations. So VVM’s first job is to investigate the combinations of effects that lead to critical situations in urban road traffic. To cover the event space, existing databases and expert knowledge are used, in addition to dedicated simulations that are being developed. As a result, the test space can be reduced to scenarios that are actually relevant, making the amount of effort involved in tests more manageable.

The second step will be to use the findings to develop a safety concept and a functional concept for describing automated systems, which can also be applied to hierarchical subsystems and components. This will make it possible, in the future, to validate new components in independent tests – instead of performing labor-intensive real test drives as is currently the case. In the final step, an example implementation of the validation framework will be produced as a demonstration. The goal is to produce an integrated, dynamic test environment in which the various test platforms can be combined flexibly from simulation right through to real drives, while at the same time allowing for an evaluation of the overall safety level. This will mean that tests can be moved systematically from the real world to simulation, resulting in greater time and cost efficiency.

Related publications:
Westhofen et al., “Criticality Metrics for Automated Driving: A Review and Suitability Analysis of the State of the Art,” https://arxiv.org/abs/2108.02403.
Neurohr et al., “Criticality Analysis for the Verification and Validation of Automated Vehicles,” in IEEE Access, vol. 9, pp. 18016–18041, 2021, doi: 10.1109/ACCESS.2021.3053159. https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9330510