An information freshness-based digital twin model to support multi-level complementary dynamic scheduling in Shared Manufacturing

Shared Manufacturing (SM) seeks to efficiently utilize idle manufacturing capacity by matching it with demand. However, disruptions in the production process are common due to the open nature of the social environment. Therefore, rescheduling becomes crucial. First, previous rescheduling efforts have primarily focused on individual aspects, overlooking integrated multilevel rescheduling. To address this gap, we propose a multilevel complementary rescheduling model that considers manufacturing services, transportation, and bin-packing problems simultaneously. This approach significantly mitigates the impact of disturbances on production. Additionally, detecting disruptions during rescheduling is essential. However, existing real-time detection systems are often impractical due to high computing resource requirements and broadband constraints. To overcome this challenge, we introduce an information freshness-based approach, leveraging digital twin technology to optimize information collection frequency within resource limitations. Furthermore, previous rescheduling strategies have typically addressed single disturbances, neglecting interrelationships among disruptions. We introduce a global customer-oriented rescheduling rule that anticipates the combined effects of multiple disturbances using digital twin-based predictions. Finally, we propose enhancements to the artificial bee colony algorithm, incorporating a self-adjusting population and a solver-based local search method to optimize scheduling. A case study in SM validates the effectiveness of our proposed approach.