Safe transportation starts with stable loading. A new study from the Faculty of Management, Economics and Social Sciences explores how real-time physical simulation and physics-informed learning can make cargo loading more reliable and significantly safer.
The research by Philipp Gabriel Mazur, Johannes Werner Melsbach, and Prof. Detlef Schoder, published in Operations Research Perspectives, presents two innovative approaches to solving the so-called Pallet Loading Problem (PLP): a dedicated real-time physics simulation and a machine-learning model informed by physical laws. Both are designed to capture the dynamic behaviour of cargo items during the loading process, a crucial factor for transport safety.
Unlike traditional static approaches that provide only snapshots of the loading state, these new methods consider how cargo physically behaves as it is placed onto pallets. This means that potential risks, such as shifting loads, damaged goods, or even environmental hazards, can be predicted and avoided more effectively.
Specifically, the research utilises a real-time physics engine similar to those used in games and digital design, and trains a recurrent neural network (LSTM) on large datasets of real-world loading sequences. Both methods outperform state-of-the-art static techniques in accuracy, although they require more computation time.
One aim of the research team was to shed light on the applicability of the different approaches for the practical scenario of air cargo palletising. The results show that in complex loading scenarios, these approaches offer substantial benefits. For researchers and practitioners alike, the future of safe and efficient loading lies in blending physical simulation with data-driven intelligence.
- To the Study: Mazur, Philipp Gabriel, Johannes Melsbach, and Detlef Schoder. “Physical Question, Virtual Answer: Optimized Real-Time Physical Simulations and Physics-Informed Learning Approaches for Cargo Loading Stability.” Operations Research Perspectives 14 (2025).
- To the webpage of the Cologne Institute for Information Systems (CIIS)