Value based resource grouping simplifies digital twin simulations
Data-Driven Resource Aggregation for Scalable Digital Twin and Simulation Model Generation
Machine Learning
Summary
Making digital twins—computer models that mimic real-world factories—can be very complicated when every single machine is shown in full detail. The authors developed a way to group similar machines together automatically by looking at how they behave during production. This makes the models simpler and faster to run while still giving accurate results. Their method uses a type of learning called inverse reinforcement learning to find which machines perform similarly. Tests show this approach keeps the model useful even when it reduces complexity a lot.
What this means in practice
- •For manufacturing system engineers: Reduce simulation complexity by grouping equivalent resources to speed up digital twin updates in dynamic factory environments.
- •For industrial automation teams: Use production event data to automate simplified resource models, easing integration of simulation into process control decisions.
Authors
Yucheng Wu, Alp Akcay
Abstract
As industrial digitization accelerates, the use of digital twins has become widespread. However, modern manufacturing systems are highly dynamic and frequently reconfigured, making manual model development and updates increasingly difficult. Recent research on automated simulation model generation shows that historical data such as event logs can yield simulation models automatically. However, modeling every resource at full granularity leads to unnecessary complexity and computational cost. We propose a method for automatically selecting and aggregating resources to simplify simulation modeling. Our method uses ideas from inverse reinforcement learning to learn behavioral value representations of resources from production event logs. This process identifies and aggregates resources with similar functional contributions. Experiments on manufacturing systems, using throughput and system time as performance measures, demonstrate that value-based resource aggregation performs well when the goal is to substantially reduce complexity while preserving performance accuracy.