The University of Southern California (USC) received a $1,500,000 Project Grant award from the National Science Foundation's (NSF) Computer and Information Science and Engineering (CISE) Federal Grant Program (CFDA 47.070). The award will fund a three-year research project titled "SLES: A Neurosymbolic Approach for Safe Multi-Agent Systems," which aims to develop new theories and algorithms for designing safe, learning-enabled multi-agent systems with applications in areas such as wildfire prevention using drone swarms and robotic systems in semi-automated warehouses. The project will leverage a combination of statistical and formal verification techniques, as well as reinforcement learning methods, to provide provable safety guarantees for these complex, high-dimensional multi-agent systems. Key deliverables include a formal specification language for multi-agent systems, robust training algorithms, and real-time algorithms for self-organizing drone swarms and warehouse robots. The research is expected to have significant societal impact and will also involve educational initiatives at the undergraduate and graduate levels at USC.
Generated 2/25/25, 4:54 AM