Detectability based prioritization of interdependent supply chain risks

Abstract

Supply chain risks must be assessed in relation to the complex interdependent interaction between these risks. Generally, risk registers are used for assessing the importance of risks that treat risks in silo and fail to capture the systemic relationships. Limited studies have focused on assessing supply chain risks within the interdependent network setting. We adapt the detectability feature from the Failure Modes and Effects Analysis (FMEA) and integrate it within the theoretically grounded framework of Bayesian Belief Networks (BBNs) for prioritizing supply chain risks. Detectability represents the effectiveness of early warning system in detecting a risk before its complete realization. We introduce two new risk measures and a process for prioritizing risks within a probabilistic network of interacting risks. We demonstrate application of our method through a simple example and compare results of different ranking schemes treating risks as independent or interdependent factors. The results clearly reveal importance of considering interdependency between risks and incorporating detectability within the modelling framework.