Automated Simulation Generation by Large Language Models

Abstract

This study introduces a new approach for automatically generating process simulations for production lines, specifically tailored to the manufacturing industry. It leverages Large Language Models (LLMs) to convert natural language inputs into a format suitable for process simulation, presenting a method and evaluation for generating efficient simulation models. The focus of this research is particularly on discrete simulation, emphasizing production line modeling in the manufacturing sector. Traditionally, a major challenge for manufacturers modeling their own factories has been the construction of simulation model structures. However,with the LLM-based approach proposed in this study, users can intuitively and efficiently build simulation models. This novel method is expected to simplify the design and execution of simulations, contributing to increased production efficiency and process optimization in the manufacturing industry.