Mitigating disruptions in a multi-echelon supply chain using adaptive ordering

Abstract

Supply chains often experience significant economic losses from disruptions such as facility breakdowns, transportation mishaps, natural calamities, and intentional attacks. To help respond and recover from a disruption, we investigate adjustments in order activity across four echelons including assembly. Simulation experiments reveal that the impact of a disruption depends on its location, with costlier and longer lasting impacts occurring from disruptions at echelons close to ultimate consumption. Cost functions based on system inventory and service can be quite ill-behaved in these complex problem settings. Expediting, an adaptive ordering approach often used to mitigate disruptions, can trigger unintended bullwhip effects, and hurt rather than help overall performance. As an alternative to expediting interventions, dynamic order-up-to policies show promise as an adaptive mitigation tool. We also find benefits in the dynamic policies from incorporating a metaheuristic parameter search over multiple echelons, yielding significantly better solution quality than embedded unimodal search.