Analyzing Service Levels in Location Problems with Disproportionate Assignment Costs

A common assumption in facility location modeling is that customer assignment cost is proportional to distance. In this work, we drop this assumption and analyze the subsequent effect on service level when weighted average... [ view full abstract ]

A common assumption in facility location modeling is that customer assignment cost is proportional to distance. In this work, we drop this assumption and analyze the subsequent effect on service level when weighted average distance is minimized. Computational experiments contrast the p-median with an augmented version enforcing higher overall service levels. We show that our modeling approach is more economical than implementing a closest assignment restriction. Additionally, a Greedy Random Adaptive Search Procedures (GRASP) algorithm is applied to solve larger problem instances.