Contributions to a Theory of Patrol Scheduling

The problem considered in this paper is that of scheduling police patrols in a random pattern. This involves generating patrol routes as well as schedules for dispatching patrol vehicles. A solution to this problem is obtained by specifying minimum average patrol requirements on each route segment in a network and then developing a procedure which meets these requirements while minimizing the total patrol effort. Introducing vehicles into the network in a Poisson stream results in Poisson streams in each route segment and so ensures that an observer cannot use previous history for predicting arrival patterns. This solution also has the property that the number of patrol cars in the network is a Poisson random variable for which the steady-state can be achieved immediately. The steady-state distribution function is also used to determine the number of patrol cars required.