Autonomous systems in general and self-driving cars in particular, hold the promise to be one of the most disruptive technologies emerging in recent years. However, the safety and resilience of these systems, if not proactively addressed, will pose a significant threat potentially impairing our relationship with these technologies and may lead to a societal rejection of adopting them permanently. This project seeks to address such concerns by equipping autonomous systems with an additional layer of intelligence allowing them to be resilient-by-cognition.
The research addresses resilience for autonomous cyber-physical systems (CPS) by integrating concepts from game theory, formal methods, and controls. Our proposed approach includes: (i) a principled framework for formally reasoning about cognitive CPS; that is, given a set of strategies captured in a formal language (e.g., temporal logic), the proposed framework builds on ideas from evolutionary game theory to understand which strategies lead to the best fit when operating in adversarial environments (ii) On-the-fly, correct-by-design feedback controller synthesis that executes the chosen strategy while satisfying physical constraints imposed by the micro-dynamics of the underlying CPS (iii) a data-driven strategy-mining approach that addresses the fundamental problem of designing the library of strategies from human demonstrations. We will illustrate our approach over key applications including self-driving cars and autonomous drone swarms. Our educational plan engages not only graduate students but also high school and undergraduate students. It also reaches out to engineers and the lay public, by providing open source implementations of our algorithms making them available both to industry and independent developers.
This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.
