Stochastic optimization of supply chain resilience under ripple effect: A COVID-19 pandemic related study

This paper presents a multi-portfolio approach and scenario-based stochastic MIP (mixed integer programming) models for optimization of supply chain operations under ripple effect. The ripple effect is caused by regional pandemic disruption risks propagated from a single primary source region and triggering delayed regional disruptions of different durations in other regions. The propagated regional disruption risks are assumed to impact both primary and backup suppliers of parts, OEM (Original Equipment Manufacturer) assembly plants as well as market demand. As a result, simultaneous disruptions in supply, demand and logistics across the entire supply chain is observed. The mitigation and recovery decisions made to improve the supply chain resilience include pre-positioning of RMI (Risk Mitigation Inventory) of parts at OEM plants and ordering recovery supplies from backup suppliers of parts, located outside the primary source region. The decisions are spatiotemporally integrated. The pre-positioning of RMI implemented before a disruptive event is optimized simultaneously with the RMI usage and recovery supply portfolios for the backup suppliers in the aftermath periods. The recovery supplies of parts and production at OEM plants, are coordinated under random availability of suppliers and plants and random market demand. The resilient solutions for the resilient supply portfolios are compared with the non-resilient solutions with no recovery resources available. The findings indicate that the resilient measures commonly used to mitigate the impacts of region-specific disruptions can be successfully applied for mitigation the impacts of multi-regional pandemic disruptions and the ripple effect.