Trustworthy CPS

Cyber-physical systems (CPS) integrate physical systems with sensors, actuators and embedded computational devices with networking capabilities, enabling them to communicate among themselves and with humans over the internet. These systems are safety- and security-critical because any failure or tampering attempt can seriously threaten human lives or the environment. Over the past two decades, advancements in high-performance, cost-effective embedded computing have transformed CPS from resource-constrained microcontroller-based systems into intelligent, autonomous entities. The increasing availability of powerful embedded devices capable of running advanced artificial intelligence (AI) software has shifted the paradigm from optimising hardware constraints to enhancing real-world systems with perception, reasoning, and decision-making capabilities. Thus, CPS can nowadays execute tasks autonomously without or with limited human supervision. The research focus in Trustworthy Autonomous CPS consists in ensuring that CPS can operate reliably and safely under various conditions while protecting them against cyber threats. In particular, at TU Wien, we develop formal methods, resilient self-adaptive mechanisms, and human-in-the-loop approaches to improve trust in autonomous CPS. Applications in autonomous vehicles, robotics, smart grids, smart cities, industrial automation, and healthcare highlight the need for rigorous testing and efficient verification methods.