A Framework for Virtual Testing of ADAS

Virtual testing of advanced driver assistance systems (ADAS) using a simulation environment provides great potential in reducing real world testing and therefore currently much effort is spent on the development of such tools. This work proposes a simulation and hardware-in-the-loop (HIL) framework, which helps to create a virtual test environment for ADAS based on real world test drive. The idea is to reproduce environmental conditions obtained on a test drive within a simulation environment. For this purpose, a production standard BMW 320d is equipped with a radar sensor to capture surrounding traffic objects and used as vehicle for test drives. Post processing of recorded GPS raw data from the navigation system using an open source map service and the radar data allows an exact reproduction of the driven road including other traffic participants. The proposed framework automatically generates the recorded and post processed environment within the simulation and enables testing of ADAS strategies by identification and implementation of a detailed vehicle model of the test vehicle. Further, connection of the simulation environment to an engine test bench allows for HIL testing and additional performance evaluation of an ADAS by e.g. measuring engine emissions or fuel consumption. Additionally, the approach easily allows extensions in terms of sensor stimulation and automatic traffic object generation based on real world measurements.