A Districting and Location Model Using Spatial Queues

Planning of efficient emergency response systems is of critical nature to modern societies, especially in densely populated areas where incidents are more likely to occur. In this paper, a model is developed for optimizing the location of an emergency response system, taking into account the stochastic nature of the network served. The model examines the problems of districting and dispatching in an integrated manner and links them to the location problem in an explicit effort to make the emergency response system as efficient as possible. The system's servers are treated as spatially distributed queues with different service rates. The model is hybrid, using a Maximal Coverage Location Problem (MCLP) formulation to restrict possible server positions with a minimum coverage level and an extended Hypercube model for districting and dispatching purposes. The model's objective is to find server locations so that the mean response time is minimized for a pre-specified minimum level of coverage. The model is applied to the Athens Urban Transport Authority for locating their tow-away vehicles throughout the Athens bus network during the Summer 2004 Olympics. TRB 2004 Annual Meeting CD-ROM Paper revised from original submittal.