LLM-driven multi-agent collaborative robotic framework

The convergence of 6G communication, LLM, and embodied agents paves the way for revolutionary advancements in multi-agent collaborative robotics. This thesis introduces an LLM-driven multi-agent collaborative robotic framework that harnesses the power of 6G technology to orchestrate intelligent swarms of robots working collectively towards complex tasks. The framework leverages LLMs' natural language processing capabilities to enable seamless human-robot interaction, facilitate robust communication between collaborating robots, and dynamically adapt to unforeseen situations. The high bandwidth and ultra-low latency of 6G networks empower real-time communication, information sharing, and coordinated decision-making within the robot swarm. This thesis explores the design and implementation of the framework, addressing key challenges such as task decomposition and plan generation, inter-agent communication and coordination and dynamic decision-making and adaptation. The thesis evaluates the framework's performance through simulations and real-world experiments, demonstrating its effectiveness in various collaborative robotic scenarios. This research contributes significantly to the field of multi-agent collaboration by introducing a novel framework that leverages the combined strengths of LLMs, 6G communication, and embodied agents. It unlocks exciting possibilities for the future of robotics, promising a future where robots work seamlessly alongside humans to solve complex challenges in diverse domains.