Multi Agent Robotic System

Multi-agent robotic systems (MARS) research focuses on designing and controlling groups of robots to collaboratively achieve complex tasks. Current efforts concentrate on improving robustness and scalability through distributed optimization techniques like ADMM, incorporating advanced learning methods such as reinforcement learning and large language models (LLMs) for high-level planning and human-robot interaction, and developing safe and efficient control strategies that address challenges like deadlocks and connectivity constraints. These advancements hold significant promise for revolutionizing various fields, including construction, logistics, and exploration, by enabling more efficient and adaptable automation in complex and uncertain environments.