Für Studenten

Für Unternehmen






RWTH Aachen

Improving the Use of Automated Simulations for Intelligent Vehicles

Assistant Professor Dr. Christian Berger

Today’s vehicular functions already depend heavily on data perceived from the surroundings. By understanding a scene, these functions support the driver in tedious or dangerous traffic situations. Thus, today’s vehicles are able to park automatically or brake fully autonomously if a crash is unavoidable for example. Even more, by the end of this decade, some automotive manufacturers announced to provide self-driving cars. While safety functions can be tested on proving grounds with specific experimental setups, autonomously driving vehicles need to act in a huge variety of different traffic situations. Thus, it is not practical to validate the algorithms of such vehicles by conducting large field tests only. Therefore, simulations complement tests on proving grounds and extensive field tests at earlier development stages. However, running simulations every time for any changes in the source code is time-consuming and inefficient. In this talk, approaches will be presented how to address these challenges.