Surveillance Camera Location Models on a Public Transportation Network

This article studies the problem of locating surveillance cameras in the context of a public transportation system. A network of stops or stations is considered which is interconnected by a set of predetermined bus routes. The studied problem is to choose the set of stations to be monitored by cameras in order to simultaneously optimize two objectives: the expected number of crimes detected by the cameras, and the image quality of the entire surveillance system. Two mathematical models based on integer programming are proposed for this problem, considering multiple periods, budget constraints, and connectivity constraints which ensure that at least a surveillance camera is assigned to one station for each pair of directly connected stations. A comparison of the performance of the proposed mathematical models using a commercial optimizer is performed using a set of randomly generated instances with 20-200 stations. The computational results show the capability of the proposed mathematical models to find optimal solutions and the required computational resources