Abstract
Utility companies typically send their meter readers out each day of the billing cycle in order to determine each customer's usage for the period. Customer churn requires the utility company to periodically remove some customer locations from its meter‐reading routes. On the other hand, the addition of new customers and locations requires the utility company to add new stops to the existing routes. A utility that does not adjust its meter‐reading routes over time can find itself with inefficient routes and, subsequently, higher meter‐reading costs. Furthermore, the utility can end up with certain billing days that require substantially larger meter‐reading resources than others. However, remedying this problem is not as simple as it may initially seem. Certain regulatory and customer service considerations can prevent the utility from shifting a customer's billing day by more than a few days in either direction. Thus, the problem of reducing the meter‐reading costs and balancing the workload can become quite difficult. We describe this Balanced Billing Cycle Vehicle Routing Problem in more detail and develop an algorithm for providing solutions to a slightly simplified version of the problem. Our algorithm uses a combination of heuristics and integer programming via a three‐stage algorithm. We discuss the performance of our procedure on a real‐world data set. © 2009 Wiley Periodicals, Inc. NETWORKS, 2009