S4OS: Scalable analysis and Synthesis of Safe, Small, Secure and Optimal Strategies for Cyber-Physical Systems

Cyber-Physical Systems (CPS) are emerging in almost every area of modern society, from intelligent transport, smart energy, smart cities, to smart health care. The rapid growth of machine-learning techniques in CPS leads to better products in terms of performance, efficiency and usability. However, CPSs are often safety critical (e.g., self-driving cars and medical devices), and the need for verification against potential fatal accidents and security attacks is self-evident and of key importance.

S4OS will develop a new generation of mathematically well-founded scalable methods and tools, integrating machine learning and model-based verification techniques in a unified framework for constructing optimal cyber-physical systems that are guaranteed to satisfy crucial safety and security constraints, and can be certified to do so.