Chuiyuan: A Large Language Model‐Driven AI Agent for Plant Factories‐Applications and Perspectives

ABSTRACT

As a vital production carrier of efficient and intensive modern agriculture, the plant factories with artificial lighting (PFALs) represents an important application scenario for the development of smart agriculture. To achieve autonomy and intelligence in environmental control, crop growth management and production decision‐making in PFALs, we developed an artificial intelligence agent (AI Agent), Chuiyuan (垂元), tailored for PFALs scenarios. Chuiyuan integrates the capabilities of large language models (LLMs) in semantic understanding, logical reasoning and interaction, and the technical advantages of deep learning in feature extraction and predictive modelling. Through LLMs, knowledge interaction in PFALs is realized, endowing Chuiyuan with professional consultation and continuous iterative optimization capabilities. In a comparative evaluation involving 104 PFAL‐related questions, Chuiyuan achieved a high score compared to a general large model. This paper further discusses the development paths and potential values of future research on PFALs AI Agent. AI Agents are expected to enhance operational efficiency, resource utilization, and production stability in PFALs. It provides a scalable intelligent decision‐making framework for digital transformation and smart agriculture applications, and offers meaningful insights for advancing controlled‐environment agriculture towards high‐efficiency, precise, and unmanned operation.