Premium
An ABC heuristic for optimizing moveable ambulance station location and vehicle repositioning for the city of São Paulo
Author(s) -
Andrade Luiz Augusto C. G.,
Cunha Claudio B.
Publication year - 2015
Publication title -
international transactions in operational research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.032
H-Index - 52
eISSN - 1475-3995
pISSN - 0969-6016
DOI - 10.1111/itor.12160
Subject(s) - context (archaeology) , ambulance service , order (exchange) , computer science , service (business) , heuristic , point (geometry) , operations research , location model , business , facility location problem , operations management , geography , medical emergency , finance , marketing , engineering , mathematics , medicine , artificial intelligence , geometry , archaeology
Abstract In this paper, we address the problem of determining the optimal location of ambulance stations, as well as the vehicle allocation and repositioning for the mobile emergency care service of São Paulo (SAMU‐SP), in Brazil. This problem arises in the context of seeking to reduce expected ambulance response times, which was within 27 minutes in São Paulo for 98% of the requests. In order to bring down total response times closer to internationally acceptable standards, SAMU‐SP devised the concept of moveable ambulance stations that can be installed in available public spaces, such as squares and parks, and can be periodically relocated to ensure a good coverage at all times. This new concept, however, was not an easy sell. It was necessary to demonstrate clearly the benefits that such stations, if properly located, could provide in the context of limited budgetary resources when compared to the traditional facilities in regular buildings. In this context, we propose a novel artificial bee colony (ABC) algorithm to guide SAMU‐SP in its strategic decisions involving their service network, as well as in the allocation and repositioning of ambulances to each stand‐by point in order to cope with varying demand at different time periods. This model was applied to analyze different scenarios, including one that was implemented in the short term and yielded an improvement of over 40% in the expected coverage.