Test Data Management

Impact on Homologation and Type Approval Process

Parallel to the shift to data-driven development and software-defined vehicles, the approval and homologation of automated driving functions is also changing. In the past, homologation was the last step during or after development. Here, subsets of tests were undertaken representing the identified hazards at vehicle level to check the overall system behavior. Due to the complexity of the driving task and subsequently of the systems, however, this is no longer sufficient, as the residual risk of such a highly automated system can no longer be determined at the overall vehicle level.

The reasons why comprehensible and uniform test data management is needed have become sufficiently clear. On the one hand, test strategies are becoming more complex and holistic, and on the other hand, the data required to correctly evaluate the results of the test strategy are more diverse.

For a comparable reason, type approval and homologation are also more complex and must rely on test or analysis results achieved based on the test data management, and on the proper implementation of the test data management.

The homologation of automated driving functions aims at proving the safety of the driving functions, as already presented in chapter 5.1. These criteria include proof that the driving function performs the driving task independently within the respective defined operating range, and that it complies with traffic regulations and it is able to autonomously switch into minimum risk conditions. As the test strategies described in this report represent one pillar necessary to approve all this, it is at the same time necessary to make this evaluable and assessable so that technical services can audit this. Consistent data management is indispensable for this. This report describes one possible approach.