OSE3AD - Open Simulation Environment for the Exploration and Evaluation of Autonomous Driving Functions

Autonomous driving (“AD”) – especially in the context of urban mobility – is one of the main trends in the automotive domain. AD is also expected to impact various social challenges such as mobility behavior in general, safety, emission and air quality, traffic congestion, as well as social inclusion (“mobility-for-all“). With AD application, the complexity for design and validation of highly automated driving functions is growing; it is obvious that the validation can no longer be performed solely on a real vehicle. For this reason, a new virtual approach is required relying on simulation driven, rule-based scenario evaluation and automated test execution. For the specific use case of urban busses for people transport which is organized around the three following main objectives: Obj 1.: Provide an open and integrated simulation framework for the efficient development, testing and validation of innovative autonomous driving functions to increase efficiency through frontloading and to enable virtual validation. Obj 2.: Identify objective evaluation criteria with respect to the selected use case of urban busses to assess both quantitative and qualitative targets for rule-based reasoning of complex scenarios, thus enabling automated scenario evaluation. Obj 3.: Design, evaluate and validate innovative autonomous driving functions for SAE level 4 using the provided simulation framework to demonstrate the increased efficiency. The OSE³AD project has an enabling character to deepen business relationships between Austria and Singapore for exploitation of autonomous driving vehicles used in public and private traffic. Such collaboration will also open the door for future application of the resulting procedures worldwide, since Singapore is a globally leading location in view of implementing autonomous driving in urban traffic. While current European activities focus on AD functions for highway driving, urban scenarios are far more complex to realize and thus go substantially beyond the state of the art.