Construction of resilient and sustainable supply chain based on multilayer Bayesian network

Resilience to disruptions is important to supply chains, while sustainability is another topic that has received widespread attention, and there are interactions and even conflicts between the two. To cope with the complexity of constructing resilient and sustainable supply chains, a new multilayer Bayesian network is developed in this paper, which provides effective support for portraying causality in supply chain networks as well as risk inference. Based on the proposed multilayer Bayesian network, the causal relationship between triggers, risk events, and risk consequences are captured, and risk triggers are evaluated from both risk probability and risk damage. These results are integrated as inputs into the process of building resilient and sustainable supply chains. Empirical analyses demonstrate that the effectiveness and great potential of the proposed model. Furthermore, delivery reliability, hurricane, and net working capital are identified as the most critical triggers. Implementing specific interventions for different triggers can significantly reduce the overall cost of constructing supply chains. This study not only offers an effective construction model but also establishes a new framework for risk analysis and control.