LLM-Powered Operator Advisor for Human-Robot Collaboration in Sustainable Assembly-Disassembly Cells

Abstract

In response to the manufacturing shift toward customization and human-centered approaches in Industry 5.0, this paper presents a conceptual framework for a virtual operator advisor based on the Large Language Model (LLM). The LLM-powered advisor processes real-time data from the assembly or disassembly line, classifying it into operational efficiency, ergonomics, and sustainability metrics to deliver context-aware, adaptive guidance tailored to each task. It is achieved by cross-referencing these data with domain knowledge stored in the database and LLM capabilities. The Advisor recommends actions toward operations optimization, potential cognitive and physical strains, and environmental responsibility. As a framework that promotes workflow adaptability, sustainability, and worker well-being, it can contribute to the evolution of human-robot collaboration in modern manufacturing.