What network and performance indicators can tell us about supply chain and sourcing resilience (and what they cannot)
本研究结合网络科学与离散事件仿真，分析网络指标（节点度）和绩效指标（准时交付、履约率、恢复时间）在评估供应链韧性时的互补作用，并测试三种不同灵活度的采购策略，为决策者提供韧性评估的实用指南。
Supply chain (SC) resilience takes network connectivity and performance persistence perspectives, which supplement each other. The extant literature has developed a large body of knowledge about SC resilience's network and performance indicators. However, we are unaware of any published research combining these two perspectives in resilience assessment. Therefore, this study aims to advance our understanding of how network and performance indicators can mutually enhance each other when analysing SC resilience as both a system property (quality) and an outcome (quantity). The unique contribution of our study is a combined use of network science and discrete-event simulation allowing for mixed-method grounded integration of static and dynamic views of supply chain resilience. Using node degrees as network indicators and on-time delivery, fulfilment rate, and time-to-recovery as performance indicators, we examine reactions of these indicators to a disruption to the sourcing strategies of three different flexibility degrees. We observe that network science methods can be used to identify disruption existence while simulation methods allow quantifying performance impact. We show how and when the combined application of network and performance indicators can inform decision-makers about SC resilience, and propose a generalised guideline for a practical implementation of the developed approach. Our main conclusion is that SC resilience-assessment models can be mutually enhanced by including network characteristics and process dynamics through a combination of network analysis and simulation. • We examine a combination of both network and performance indicators in resilience analysis. • Sourcing strategies of three different flexibility degrees are tested. • We show what information each of the indicators can provide for resilience analysis. • We consider both network and process dynamics in resilience assessment.