Models and algorithms for workforce scheduling and routing problems in emergency response services

• Main Author: Xie, Fulin
• Other Authors: Potts, Christopher ; Bektas, Tolga
• Published: University of Southampton 2018
• Subjects: 510
• Online Access: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.749821

Emergency response services play a key role in protecting public safety and health, and therefore developing effective and efficient response systems is of critical importance. In this thesis, we focus on the workforce scheduling and routing problems (WSRPs) that are commonly faced by emergency response organisations. We first present a simulation model for real-time emergency vehicle dispatching and routing, developed based on a case study of a British company providing emergency road services. The developed model is used to evaluate system performance, test scenarios and compare the effectiveness of different dispatching policies. The results of simulation study motivate us to design more advanced heuristic algorithms for the static WSRP. To this purpose, we develop a simple and fast algorithm based on the iterated local search (ILS) framework. The performance of the proposed algorithm is evaluated on benchmark instances against an off-the-shelf optimizer and an existing adaptive large neighbourhood search algorithm. The proposed ILS algorithm is also applied to solve the skill vehicle routing problem, which can be viewed as a special case of the WSRP. To further improve the decision making, we exploit the stochastic information about future requests and integrate this part of information into the solution method for the dynamic WSRP. A stochastic set-partitioning model is described and integrated with a sampling-based approach. The proposed model uses a two-stage framework, where the first-stage is concerned with finding a set of feasible routes covering known requests, while the second-stage estimates the effect of the same routes with respect to future requests. The performance of the proposed model is evaluated against a deterministic model and a naive greedy heuristic within a simulation framework.