This study aims to evaluate the quality of the solutions for facility location-allocation problems generated by a GIS-T (Geographic Information System for Transportation) software. These solutions were obtained from combining the Facility Location and Transportation Problem routines, when compared with the optimal solutions, which were obtained using the exact mathematical model based on the Mixed Integer Linear Programming (MILP) developed externally to the GIS. The models were applied to three simulations: the first one proposes set up businesses and customers' allocation in the state of São Paulo; the second involves a wholesaler and an investigation of distribution center location and retailers' allocation; and the third one locates day-care centers in an urban context allocating the demand. The results showed that when the facility capacity is considered, in addition to determine different locations for the new facilities, the optimal MILP model can produce results that are 37% better than those of GIS.
Geographic information system; Location-allocation problems; Transportation planning; Mixed integer linear programming
