Dynamic structural adaptation for building viable supply chains under super disruption events

Supply chain (SC) has been increasingly challenged by disruption events (DEs), where super DEs (SDEs) comprising a sequence of DEs, e.g., COVID-19, pose significant threats with longterm impacts. To hedge against SDEs, SC viability has been introduced, whose distinctive feature is the ability to adapt the SC structure. Building SC viability via dynamic SC structural adaptation under SDEs, however, has not been quantitatively addressed in the literature. This study investigates a novel viable SC building problem under SDEs. It consists of timely assessing the disruption risk and dynamically adapting the SC structure, to satisfy uncertain demands. The aim is to find the best balance between the disruption risk and the SC operational cost. To portray the structural and temporal risk propagations along the dynamic SC structure, a new structure-variable dynamic Bayesian network (SVDBN) is proposed. Then, a bi-objective mixed- integer non-linear programming (MI-NLP) model is established. Based on analyses of problem features, a decomposition-and-clustering (DC) heuristic algorithm is designed to solve the problem. Numerical experiments are conducted to evaluate the performance of the approach, and managerial insights are provided as well.