July 9, 2025

Safe mobility through AI - District Administrator visits DLR Sankt Augustin

  • District Administrator Sebastian Schuster (Rhein-Sieg-Kreis) learnt about AI safety and security research at the DLR site in Sankt Augustin
  • OneVision simulation platform enables research into critical handover scenarios between autonomous vehicles and human drivers
  • Emotion recognition through AI-supported camera analysis identifies excessive demands and stress in drivers
  • Data protection and social acceptance are at the centre of technology development

Visit to the Institute for AI Safety and Security

Sebastian Schuster, District Administrator of the Rhein-Sieg district, visited the DLR institutes at the Sankt Augustin site together with Regina Rosenstock, Head of the Department for Economic Development and Strategic District Development. After an introduction by site manager Dr Melanie von der Wiesche, the guests were given a detailed insight into the research work of the three institutes based here.

OneVision is a vehicle simulator for AI safety and security research

Dr Arne Raulf, Dr Markus Lange and Dr Elena Hoemann from the DLR Institute for AI Safety and Security presented the immersive simulation platform OneVision. This specially developed vehicle simulator generates targeted research data on AI safety and focusses on the camera-supported recording of the emotional state of drivers using artificial intelligence.

Researching critical handover scenarios

The research centres on a central question of future mobility: is a driver stressed or overwhelmed and therefore no longer fit to drive? Can an automated vehicle be safely handed over to a human driver? OneVision makes it possible to realistically analyse these critical handover scenarios between autonomous AI systems and human operators.
The high reliability of the system with minimal sensor input is particularly impressive - a single camera is sufficient to capture the emotional state. This is only possible by using AI to analyse the data, whereas alternative solutions rely on multi-camera systems.

Focus on safety and data protection

Autonomous driving functions can be activated if the driver is recognised as being overstretched or in safety-critical situations. Iterative development processes ensure that the systems are continuously improved. At the same time, the system demonstrates scenarios for the protection of personal data from biometrics and AI - an essential aspect for the social acceptance of such technologies.

The constructive exchange with the representatives of the Rhein-Sieg district emphasises the importance of cooperation between research and regional administration for the integration of secure AI technologies into future mobility.