Next-Gen Assembly Assistance: An AI-Driven System for Industry 5.0

Manual assembly processes in small and medium-sized enterprises (SMEs) often lack the flexibility and adaptability of automated systems, which can limit productivity and increase error rates. To address these challenges, we propose an AI-based assembly assistance system that leverages Large Language Models (LLMs) to automatically generate and update work instructions, safety guidelines, and task-specific documentation in real-time. This approach offers a novel solution to the rigidity of existing assistance systems by integrating dynamic, context-aware support directly into the production environment. This study focuses on the preliminary tests conducted as part of the preparatory phase for developing the proposed system. These tests involved the generation of assembly instructions using LLMs and a systematic analysis of the resulting outputs. Key findings of this study include insights into how to optimize data input to achieve the most accurate and contextually relevant results when generating assembly instructions. Furthermore, the analysis identifies the most common errors encountered in the automated generation of assembly instructions, providing a deeper understanding of the limitations and areas for improvement in the system’s performance. This research contributes to the development of flexible, human-centered assistance technologies, paving the way for the future of AI-supported manual assembly.