Identification of real world driving scenarios for the functional safety of autonomous vehicles

摘要

The knowledge of the real world driving environment for current road vehicles is fundamental in order to better identify safety problems encountered by autonomous vehicles in similar complex situations. This paper presents the context and the methodology applied to collect data, extract the real world driving scenarios and statistically identify the variables playing a role in such scenarios. Thus, it will be possible to improve the robustness of the architectures and systems of the future autonomous vehicles being developed at the VEDECOM Institute.