ChatEMT: large language model framework for energy-efficient machine tools – review, paradigm, and perspectives

The increasing demand for sustainable manufacturing has intensified the need for energy-efficient machining solutions that support economic, environmental, and societal sustainability. While emerging large language models (LLMs) have been successfully applied to other manufacturing problems, their application to energy-efficient machine tool management remains limited, leaving significant gaps in integrating LLM intelligence with energy-oriented machining services. To address this challenge, this paper proposes a novel LLM-enabled framework for energy-efficient machine tools, termed ChatEMT. This framework provides explainable, user-friendly, and semantically grounded energy services (e.g. analysis, monitoring, and optimisation) for sustainable machining. First, recent studies on the energy consumption of machine tools are systematically reviewed. Then, the paradigm and architecture of ChatEMT are presented, in which LLMs serve as the core layer that bridges energy data, domain knowledge, and diverse models. Specifically, LLMs incorporate domain knowledge of machine tool energy, retrieve production information, perform semantic reasoning over machining processes, and orchestrate tool invocations to support a wide range of energy services. Finally, open challenges and future research directions are discussed to guide subsequent studies. This work highlights the critical role of LLMs in enabling explainable, flexible, and context-aware energy intelligence for machining systems.