Designing an LLM-Based Copilot for Manufacturing Equipment Selection
Jonas Werheid, Oleksandr Melnychuk, Hans Zhou, Meike Huber, Christoph Rippe, Dominik Joosten, Zozan Keskin, Max Wittstamm, Sathya Subramani, Benny Drescher, Amon Göppert, Anas Abdelrazeq, Robert H. Schmitt
TL;DR
The paper tackles the challenge of accelerating ramp-up in manufacturing by introducing a factual-driven copilot that combines large-language models with Retrieval-Augmented Generation in a state-machine framework. It targets equipment selection for robots, feeders, and vision systems, leveraging both relational and semi-structured knowledge to produce structured, traceable recommendations. Industrial evaluation shows the system correctly suggests equipment in most prompts (19/22) with full requirement satisfaction in 6 cases, demonstrating practical potential and highlighting gaps in layout design and ramp-up deployment. The approach promises to reduce ramp-up time and improve decision quality, with future work aimed at end-to-end integration from design through ramp-up implementation.
Abstract
Effective decision-making in automation equipment selection is critical for reducing ramp-up time and maintaining production quality, especially in the face of increasing product variation and market demands. However, limited expertise and resource constraints often result in inefficiencies during the ramp-up phase when new products are integrated into production lines. Existing methods often lack structured and tailored solutions to support automation engineers in reducing ramp-up time, leading to compromises in quality. This research investigates whether large-language models (LLMs), combined with Retrieval-Augmented Generation (RAG), can assist in streamlining equipment selection in ramp-up planning. We propose a factual-driven copilot integrating LLMs with structured and semi-structured knowledge retrieval for three component types (robots, feeders and vision systems), providing a guided and traceable state-machine process for decision-making in automation equipment selection. The system was demonstrated to an industrial partner, who tested it on three internal use-cases. Their feedback affirmed its capability to provide logical and actionable recommendations for automation equipment. More specifically, among 22 equipment prompts analyzed, 19 involved selecting the correct equipment while considering most requirements, and in 6 cases, all requirements were fully met.