A project to establish generally accepted quality criteria, tools and methods as well as scenarios and situations for the approval of highly automated driving functions.
The purpose of the PEGASUS project was to develop common methods and tools for the scenario-based testing of highly automated vehicles in order to counter technical, legal and social challenges to be encountered on their approval.
BMWi (Federal Ministry for Economic Affairs and Energy)
- TraceTronic GmbH
- Adam Opel AG
- AUDI AG
- BMW AG
- Daimler AG
- Volkswagen AG
- A.D.C. Automotive Distance Control Systems GmbH
- Continental Teves AG
- Robert Bosch GmbH
- TÜV SÜD GmbH
- Forschungsgesellschaft Kraftfahrwesen mbH Aachen
- iMAR Navigation GmbH
- IPG Automotive GmbH
- QTronic GmbH
- VIRES Simulationstechnologie GmbH
- DLR e.V. – Institut für Verkehrssystemtechnik
- TU Darmstadt – Fachgebiet Fahrzeugtechnik
January 2016 to June 2019
Automated driving is a key issue in the development of the automotive sector. The highest quality and safety requirements exist in relation to the approval of highly automated vehicle functions. Present procedure models and testing methods are not adequate for this purpose. In order to automatically check compliance with these requirements, there is a great need for appropriate quality criteria, validation methods and software tools.
The purpose of the PEGASUS project was to develop generally accepted methods and tools for testing highly automated vehicle functions. This prepares the approval of highly automated vehicles for the market. To this end, the first step is to conduct preliminary work to determine the performance level of HAD functions (e.g. the definition of an "average driver" as a benchmark), from which criteria relating to the standard and quality can be derived. Once these were in place, testing methods, testing catalogs as well as testing instruments for simulation, laboratory and test bed-based concepts were developed in the project. PEGASUS was thereby intended to provide a reference for the efficient validation of highly automated driving functions.
Within the framework of this project, we developed methods and software components for tool chains to validate highly automated driving functions. For the resulting requirements, new methods are developed for use in test automation, trace analysis and test evaluation and then realized as software demonstrators. In order to take account of the large number of variants and the high safety requirements on highly automated vehicle functions, existing trace analysis approaches are continued and new methods created for the automated evaluation of the large and complex quantities of simulation and test driving data thereby obtained.
On completion of the project, we will use the results to expand our existing product portfolio to include new features, libraries and tools for the testing and safeguarding of highly automated vehicle functions. The solutions will be integrated into the tool chains of our project partners for use there.