Resilience modeling of supply chain partners using Bayesian networks

Abstract

Supply chains are increasingly exposed to disruptions caused by natural disasters, technical failures, and human error, with these vulnerabilities being amplified by globalization and growing system complexity. These challenges have highlighted the need for resilient supply chain (SC) designs that can withstand disruptions, sustain operational performance, and recover efficiently. Although previous research has predominantly examined supply chain resilience (SCR) from a qualitative perspective, quantitative models that integrate both proactive and reactive decision-making remain limited. In particular, the causal interdependencies among suppliers, buyers, and transportation nodes are often overlooked. This paper addresses this gap by proposing a Bayesian Network (BN) framework for modeling the resilience of supply chain partners. The proposed approach captures the conditional dependencies among disruptions and supply chain entities, enabling decision-makers to evaluate resilience under a range of disruption scenarios. By quantifying system robustness, recovery capability, and the effectiveness of mitigation strategies, the BN model provides actionable, data-driven insights for strengthening supply chain resilience through probabilistic analysis.