The Ambulance Location Problem in Tijuana, Mexico
This work studies the ambulance location problem for the Red Cross in Tijuana, Baja California , Mexico. The solution to the ambulance location problem is to optimally locate all available ambulances within the city such that coverage of the city population is maximized and a quick response to any emergency is ensured. The problem is posed using three different coverage models; namely the Location Set Covering Model (LSCM), the Maximal Covering Location Problem (MCLP) and the Double Standard Model (DSM), also we proposed robust versions of each model, where the goal was to find a single solution that might provide optima coverage in several different scenarios. Using real-world data collected from over 44,000 emergency calls received by the Red Cross of Tijuana, several scenarios were generated that provide different perspectives of the demand throughout the city, considering such factors as the time of day, work and off-days, geographical organization and call priority. These scenarios are solved using Integer Linear Programming and the solutions are compared with the current locations used by the Red Cross. Results show that demand coverage and response times can be substantially improved without additional resources.
KeywordsInteger Linear Program Emergency Medical Service Demand Point Robust Solution Individual Solution
The authors of this work extend a special thanks to the Red Cross of Tijuana for their support and collaboration in the development of the present research and for providing the data base of EMS records. First author was supported by scholarship provided by CNBES, Mexico. Funding for this work was provided by CONACYT (Mexico) Basic Science Research Project No. 178323, DGEST and TecNM (Mexico) Research Projects 5414.14-P and 5621.15-P, and FP7-PEOPLE-2013-IRSES project ACOBSEC financed by the European Commission with contract No. 612689. Finally, we also want to thank Jose M. Camacho Avila and Oscar E. Escobar Nungaray, both of whom worked on data cleaning and preprocessing while they were students at the Instituto Tecnológico de Tijuana in 2014.
- 5.Cruz Roja: Estadisticas realizadas por la Cruz Roja Tijuana (2012)Google Scholar
- 10.Google: Google maps distance matrix api. https://developers.google.com/maps/documentation/distancematrix/ (2014). (2014. 11. 30)
- 11.Google: Google maps static maps api. https://developers.google.com/maps/documentation/staticmaps/ (2014). (2014. 11. 30)
- 12.Google: Google places api. https://developers.google.com/places/ (2014). (2014. 11. 30)
- 13.INEGI: Censo de poblacion y vivienda 2010. http://www.inegi.org.mx/est/lista_cubos/consulta.aspx?p=pob&c=1 (2010). (2015. 2. 11)
- 14.Laporte, G., Louveaux, F.V., Semet, F., Thirion, A.: Application of the double standard model for ambulance location. In: Nunen, J.A.E.E., et al. (eds.) Innovations in Distribution Logistics. Lecture Notes in Economics and Mathematical Systems, vol. 619, Chap. 12, pp. 235–249. Springer, Berlin (2009)Google Scholar
- 16.MathWorks: Hierarchical clustering. http://www.mathworks.com/help/stats/hierarchical-clustering.html (2014). (2014. 11. 23)
- 17.MathWorks: knnclassify. http://www.mathworks.com/help/bioinfo/ref/knnclassify.html (2014). (2014. 11. 23)
- 18.MathWorks: Linear programming and mixed-integer linear programming. http://www.mathworks.com/help/optim/linear-programming-and-mixed-integer-linear-programming.html (2014). (2014. 11. 23)
- 19.MathWorks: Options for linear programming. http://www.mathworks.com/help/optim/ug/intlinprog.html?requestedDomain=www.mathworks.com#inputarg_options (2016). (2016. 02. 22)
- 20.MathWorks: Options for mixed-integer linear programming. http://www.mathworks.com/help/optim/ug/intlinprog.html#inputarg_options (2016). (2016. 02. 22)
- 21.Mitchell, T.: Machine Learning. McGraw-Hill (1997)Google Scholar