Toward Embodied Intelligence-Enabled Human–Robot Symbiotic Manufacturing: A Large Language Model-Based Perspective
Wenhang Dong, Shufei Li, Pai Zheng
Abstract
Abstract Human–robot collaborative manufacturing is crucial in modern production landscapes for flexible automation. However, existing human–robot systems face several challenges, including a lack of human-centric autonomy for adjustments, limited generalization for production variants, and ineffective synchronous teamwork with feedback. Emerging large language models (LLMs) offer a viable solution. Despite the growing interest in LLMs, their deployment within the manufacturing domain remains unexplored. This article delves into the potential of LLMs for embodied intelligence-enabled human–robot symbiotic manufacturing (HRSM). HRSM is a paradigm characterized by human centricity, generalizability, and seamless integration. LLMs can facilitate human-centric interactions, generalizable collaboration, and ensure seamless execution, paving the way for realizing HRSM. Finally, the main challenges and potential opportunities are further discussed to enable the readiness toward HRSM.