Project Grant 2219488

This Project Grant from the National Science Foundation's Computer and Information Science and Engineering program, CFDA #47.070, provides $253,618 to Michigan State University from October 1, 2022 to September 30, 2025. The funding supports research into collaborative reinforcement learning approaches for multi-robot systems that ensure safety and privacy protection. The researchers will develop a model-enabled technique combining model-based safety with model-free reinforcement learning, enabling its application to safety-critical collaborative multi-robot environments. This will first address single-robot learning using deep Koopman-based safety regulation, then extend to multi-robot collective learning where privacy is preserved through dynamics-based protection of collected and shared data. The algorithms and frameworks will be evaluated through numerical simulations and experiments with connected vehicles, with results informing graduate and undergraduate courses. The funding aims to advance computer and information science and engineering through investigator-initiated research and education.