Software Tool for Evaluation of Multi-Sensor Object Tracking in ADAS systems

Nowadays, the innovations of AI and other automated decision-making software are spreading to many different areas. The automotive field in particular is rapidly shifting towards the concepts of Advanced Driver Assistance Systems (ADAS) and Obstacle Detection and Avoidance Systems (ODAS), which could bring huge benefits in the future. However, before being able to use these tools, many assurances are required regarding their functioning and safety. To this end, several control techniques exist to evaluate the performance of these software, but a reliable and repeatable method for evaluating complex scenarios and corner cases is still lacking. In this paper we propose a suite of tools for the generation and analysis of synthetic tests, aimed at evaluating and analyzing the functioning of autonomous driving systems in order to measure their effectiveness and to drive their development.