Optimizing Police Facility Locations Based on Cluster Analysis and the Maximal Covering Location Problem

Abstract

Planning and defining places for the installation of police facilities are fundamental to improving the public security service in the urban space. Geographic Information Systems connecting the spatial distribution of police occurrences, budgetary restrictions, and the maximum distance covered are state-of-the-art innovations addressing the need for preventive and responsive police management. The present work proposes a management information system to support the analysis and suggestion of potential police facility locations. The management information system is modeled using k-means for cluster analysis and the definition of candidate locations, and the maximal covering location problem is used to optimize the predefined locations. The proposed system allows the analysis of alternative locations and their impacts on public security. The application in Brazil demonstrates that it is possible to obtain an additional 22% gain in the coverage area of occurrences and an additional reduction of 920 m in terms of the average distance covered when comparing the management information system’s suggested locations to the current configuration. Thus, our assessment provides an efficient tool for supporting decisions regarding the location of police facilities and helps improve the public security service.