Objective:​ While the technical proof of concept for automated driving was shown in various projects, the existing methods and tools for the sign-off process are not suitable for the complexity of these systems and would be inefficient with regard to costs and time resources. Thus, the project PEGASUS aims to develop an effective and generally accepted procedure for the definition of design and quality criteria for highly automated vehicles. This paper focuses on a method to reduce the testing efforts for the sign-off process.

Methods:​ One element of the developed procedure is a database containing relevant traffic scenarios for the testing of highly automated vehicles. It is embedded in a circuit of recording and integrating scenarios from real-world traffic on the one hand and extracting and providing them to different suitable test environments on the other hand. This so-called circuit of relevant situations was already outlined in the 23rd ESV conference (Eckstein and Zlocki 2013), further elaborated in the meanwhile and is currently under development within PEGASUS.

Results:​ The process of recording and integrating relevant scenarios from real world measurements into the database has been implemented. Data from different sources (such as field operational test, accident databases etc.) were formatted into a standardized format to apply a common processing chain. This processing chain includes the calculation of derived signals and scenario likelihoods. Based on these steps it is possible to identify and cluster specific scenarios within bigger data sets. Afterwards, performance indicators can be calculated for characterization of scenario groups which enables to build distributions of scenario parameters (e.g. the criticality) and to derive test specifications for the database.

Conclusion:​ The dominating challenge for the implementation of automated driving is not the technical proof of concept, but the validation of these systems. Therefore, the PEGASUS project results will make a significant contribution implementation of this new technology due to the involvement of various OEMs in the project consortium and the foundation by the Federal Ministry for Economic Affairs and Energy. This paper describes the methodology and first implementation results for the database and the process chain from data collection, data storage to scenario parameterization and test specification derivation.