Resilience of intelligent cyber-physical systems of systems
Resilienz
The cyber-physical systems of the future are smart, highly networked and autonomous. They use pioneering technologies such as AI-supported decision-making processes and machine learning to enhance their perception. However, this leap in innovation brings with it considerable hurdles. The break with previous practices cancels out almost any process-centredness achieved. In order to ensure robust operation with minimal security risks despite increasing system complexity, we need to reinvent our understanding of security: Resilience is the key term that promises a solution to the rapidly growing attack surface. Resilient systems are designed to be able to react in the event of a malicious (or accidental) compromise in order to restore the maximum achievable system function. This project takes a holistic approach to achieving resilience. To this end, resilience mechanisms are implemented and integrated at both algorithmic and system level and tested with an avionics and a man-in-the-loop use case.
The Institute of Data Science contributes innovative methods for real-time monitoring and health estimation of complex decision-making systems. These combine rule-based and formal approaches, which reach their limits due to complexity, with anomaly-oriented methods and are supplemented at overall system level by approaches from secure software engineering.
