Digitalization in Production Logistics: How AI, Digital Twins, and Simulation Are Driving the Shift from Model-based to Data-driven Approaches

Abstract

The paradigm shift from model-based to data-driven approaches in production logistics is radically transforming the manufacturing landscape. This paper delves into the profound implications of this transition, emphasizing the instrumental role of simulation and digital twins. Through an exhaustive literature review, the emerging trends in data-driven approaches and the driving forces behind this change are elucidated. A comparative case study is presented, contrasting the model-based approach, which employs predefined models and principles in simulations, with the innovative data-driven approach, which utilizes real-time data and machine learning for system monitoring and predictions in production logistics. The analysis reveals the heightened efficiency, adaptability, and effectiveness offered by data-driven approach, showcasing their superiority. Additionally, the prospective roles of AI, particularly large language models like ChatGPT, in enhancing data-driven production logistics are investigated. Exploratory scenarios envision the future trajectories of simulation and digital twin applications in this rapidly evolving field. This paper provides academia and industry with a comprehensive overview of the digitalization in production logistics, emphasizing the immense promise of data-driven approach and AI.