Abstract

The academic literature on logistics network disruptions has increased sharply recently. Disruptions are random events that cause an element of a logistics network to stop functioning, either completely or partially, for a (typically random) given amount of
time. Because of today's globalized threats such as, labor disruptions or failures resulting from harsh weather conditions, there has been a renewed interest in resilient facility
location. Design of reliable logistics networks to avoid disruption can be accomplished by fortification of existing facilities and de�fining backup facilities.
In this thesis, we will look at two components of a logistics system that can be affected by a disruption: the locations of the facilities, and the routes between a customer and a facility. We study the following three designs of logistics networks under disruption: (i) Reliable Capacitated Facility Location under Disruption, (ii) Shared Capacitated Reliable Facility Location in Presence of Disruption , and (iii) Reliable Facility Location and Routing in Logistics Network in presence of disruption considering backup sharing.
A column generation approach is proposed to model and solve all three logistics problems. Results show the effectiveness of the decomposition schemes for solving exactly much larger facility location instances than in the literature. In addition, shared backup
is shown to be a very effective scheme for the design of reliable facility locations/roads.