Robotic Systems and LLM Algorithms for Transforming Manufacturing Into Smart Autonomous Ecosystem

Recent developments at the intersection of robotic systems and large language model (LLM) algorithms are paving the way for new manufacturing systems that include intelligent, autonomous ecosystems. This chapter introduces a new framework that uses LLM-powered multi-agent systems to convert traditional factory floors into thinking and responding environments. The robots perform physical actions (e.g., pick-and-place activity, welding, assembly, etc.) and are augmented with sensors and vision systems, whereas, the LLM agents are woven into the factory floor to understand natural language instructions and feedback from the robots' sensors (e.g., temperature, motion) and coordinate actions with humans, as well as other robotic agents. This chapter provides a systematic pathway from theory of multi-agent LLMs, architecture, and use scenarios to implementation strategies (robot simulation, digital twins, LLM), and evaluation metrics (completion time, flexibility, uptime).