Aiming at the structural contradiction between "increasing demand fluctuation" and "rigid solidification of supply chain" in the global footwear industry, this paper puts forward and verifies a set of collaborative simulation and dynamic optimization system of supply chain which integrates edge computing and digital twins. On the theoretical level, the triple optimization criterion of "real-time-global-adaptive" is constructed, and the digital twin model driven by mechanism-data and the edge-cloud hierarchical decision-making mechanism are established, which fills the gap in dynamic uncertainty scene modeling. On the technical level, the millisecond response is achieved by deploying lightweight sensing devices at the edge layer, and the global scheduling algorithm with carbon emission constraints is integrated at the cloud layer, forming a closed-loop architecture of "sensing-simulation-optimization-execution-collaboration". The experiment is based on the real data of three shoe enterprises and two garment factories in China. The results show that the punctuality rate of orders is increased from 76.2% to 93.5%, the inventory turnover days are shortened by 33.6%, the response to sudden demand is delayed by 96.4%, and the transportation carbon emissions are reduced by 18.2%. The toughness evaluation shows that the nodes with high redundancy can be restored within 3.2 hours, ensuring the continuous and stable supply chain. The research provides a replicable green and efficient collaborative optimization paradigm for the footwear and apparel industry, helps small and medium-sized enterprises to access at a low threshold and promotes the industry to transform into a "smart country".
Chen et al. (Sun,) studied this question.