A human-centric large language model interface for monitoring and control of autonomous mobile robots via digital twins in industry 5.0

The diversity of industrial robot manufacturers, proprietary programming languages, and graphical interfaces complicates robotic system deployment and operation. This heterogeneity highlights the need for intuitive, accessible, and human-centric interaction methods. This paper presents a framework for natural language-based monitoring and selected control of autonomous mobile robots using a Large Language Model (LLM) interface, structured API communication, and digital twin visualization. The approach enables operators to interact with robots through natural language without requiring detailed knowledge of robot-specific interfaces or technical commands. The framework is evaluated from the operator’s perspective, focusing on its ability to simplify access to robot information, reduce interaction complexity, and support selected control tasks. Digital twin integration provides additional visual feedback on robot status and its operating environment. The proposed framework contributes to the development of more transparent, accessible, and human-oriented robotic systems aligned with Industry 5.0 principles.