Distribution Network Design

Abstract

A distribution network design is of importance to systematically organize the flow of products from supply points to demand points. A distribution network is not only crucial before sales but also after sales because regardless of the initial quality of product, malfunctions occur. These malfunctions need to be repaired by replacing defective parts, which are delivered through after sales distribution network and all activities, carried out from the report of the malfunction until the product is up and running, are called after sales services. The replacement parts are called service parts or spare parts and this thesis project focuses on the distribution network design for delivery of these items. Customer surveys point out that high quality after sales services has become a criterion for selecting supplier companies. A shorter response time to failures and availability of service parts in the storage are given importance. Because after sales services became a field that companies can differentiate themselves from the competitors, practitioners have an increasing interest on establishing the optimal distribution network design to meet customer expectations while being cost-competitive.

The researchers have addressed service parts distribution with a strong focus on inventory management. However, distribution network design involves more decisions such as number of facilities, facility locations, allocation of customer demand to these facilities, and storage capacities. In addition, the managerial aspects and strategies for service parts logistics in business case studies are derived from electronic, automotive and aviation industries. Considering these, there is a knowledge gap about distribution network design for service parts in healthcare industry. This thesis project aims at filling this gap by introducing a general mathematical model for distribution network optimization based on an empirical study, conducted in General Electric Healthcare service parts delivery network.