By analysing statistical data, server-based driver assistance systems allow the recording of influencing variables that would not be possible locally in the vehicle. This includes, in addition to the investigation of historical accident data, the detailed analysis of recorded vehicle data.


Within the framework of this project, a prediction of the criticality of the ego-vehicle under consideration is to be extended to several seconds by centrally recording the influencing variables on driving safety.


  • Evaluation and preparation of available databases
  • Analysis of static risk factors via accident databases
  • Analysis of dynamic influencing variables by applying machine learning methods to recorded vehicle signals
  • Analysis of the utilisation potential