Balancing Workloads through Co-location in Covering Problems

Abstract

Total demand suitably served and facility workload balance are two important considerations in location coverage. Previous work has dealt with workload balancing issues using a number of approaches, including imposing facility capacities and the use of multiple objectives focused on workload variation. However, a facility is usually restricted to a single service unit, inconsistent with strategies that allow for increased staffing such as multiple service units in dealing with higher levels of demand. This article proposes a new bi-objective optimization model that maximizes total demand coverage and minimizes workload differences simultaneously while allowing more than one service unit to be co-located at a site. Since the proposed model is strongly NP hard, a heuristic algorithm is developed for efficient solution. The model is applied to support postal delivery service planning. Results show that the proposed model offers improved performance compared to approaches that do not permit co-location. The proposed algorithm is able to produce high-quality solutions that evenly distribute allocated service demand, and does so much faster with higher-quality solutions compared to exact solution approaches.