Resiliency of Long Chains under Plant Disruptions

Abstract

Despite being sparse and therefore economical, long chains are repeatedly observed to be almost as effective as fully flexible manufacturing systems in dealing with demand uncertainty. Motivated by a string of events that have globally impacted supply chains, our objective is to study the resiliency of long chains when there is a positive probability of plant disruptions. To achieve this, we first investigate the connection between long and open chains. Then, we propose a simple greedy algorithm for characterizing the performance of an open chain for a given scenario of plant disruptions and product demands. When such information is uncertain, we develop a pair of Markov chains for computing the load of the open chain’s dedicated and flexible arcs, respectively, which in turn enables us to have an elaborate look at the long chain’s expected performance. We further argue that the long chain is indeed a viable alternative of the full flexibility system that provides a significant improvement over the dedicated design when the demand distribution is symmetric or right-skewed and that all three designs are similarly competitive otherwise. Finally, we report a case study on a flexible configuration of cellular base stations in the People’s Republic of China.