The risk management of customized product supply chains based on digital twin technology
Xiaohui Ren, Ning Zhao
Abstract
In light of increasing uncertainties in the global market, customized product supply chains (CPSCs) are susceptible to disruptions owing to their order-driven, zero-inventory production model and the critical requirement for timely deliveries. Consequently, developing precise disruption risk prediction and intelligent adjustment mechanisms for CPSCs is strategically crucial. Current research on supply chain digital twins (DTs) primarily addresses industries characterised by standard inventory models and tends to simplify the production processes of suppliers and enterprises in DT construction. This approach fails to provide the precision needed for CPSCs. Furthermore, current high-fidelity DT modelling approaches predominantly emphasise workshop and equipment modelling, which imposes a significant burden when implemented in CPSCs. To address these challenges, a DT-based CPSC risk management framework is proposed that combines DTs with discrete event simulation technology for disruption risk prediction and intelligent adjustment. This framework employs a multi-granularity differentiated DT modelling technique that categorises the supplier model into three distinct types for coarse-, medium-, and fine-grained modelling, ensuring prediction accuracy while reducing modelling efforts. Practical implementation in enterprise projects demonstrated an average prediction accuracy of 87.56%. Moreover, when disruption risks are predicted, the framework makes effective supply plan adjustments possible so that CPSC operational stability can be maintained.