Goal programming applied to police patrol allocation

This paper describes the development of a decision-making methodology to aid the police patrol planner in resolving the patrol allocation problem. The methodology developed provides information to assist in two major decision areas: staff sizing and vehicle requirement schedules. This research uses a queuing based simulation model to derive expected service levels for a number of different performance measures. Given these results, tradeoff curves are mapped between specified performance measures and utilization rates of patrol vehicles. With this information the patrol planner specifies limiting levels for both performance measures and other managerial constraints which, in turn, are inputted into a goal programming algorithm to determine appropriate staff sizes and vehicle requirements. The primary output provided the patrol planner is the required amount of patrol vehicle hours to allocate to each watch per day per precinct in order to satisfy a designated set of performance measure levels and managerial constraints. The methodology presented in this research was developed in conjunction with the Columbus, Ohio, Police Department with data being provided by the Research and Planning Division.