Current collaboration between the logistics and manufacturing industries faces challenges such as high barriers to multi-source heterogeneous data exchange, insufficient real-time production-distribution scheduling coordination, and low cross-system resource utilization. This paper proposes an innovative model for deep integration based on the Industrial Internet of Things (IIoT) and a cloud-edge-device collaborative computing architecture. A unified cyber-physical system (CPS) communication framework is constructed using OPC UA and MQTT protocols to achieve data synchronization between the Manufacturing Execution System (MES), Warehouse Management System (WMS), and Transportation Management System (TMS). A distributed event-driven architecture (EDA) enables real-time responses to events such as production status changes and logistics equipment status. A multi-agent deep reinforcement learning algorithm (ST-MADRL) under spatiotemporal constraints is employed to dynamically coordinate and optimize production line work orders and delivery routes. Experimental validation at a large high-end equipment manufacturing park demonstrates significant improvements in system performance: data exchange latency is reduced to less than 50 milliseconds; average order response time is shortened from 158.2 minutes to 23.3 minutes; idle rates of warehousing and transportation resources are reduced, providing a feasible technical path and implementation framework for the deep integration of the two industries.
Xiaowei Chen (Thu,) studied this question.