Tailored Adaptive Large Neighborhood Search to optimize a real-world logistic service for dependent patients

In recent years, the number of disabled people have been growing drastically. This trend makes the existing healthcare logistic services inadequate to meet patients demand and, thus, a model to improve the efficiency of this service is necessary. To fulfil this need, the goal of this research is the development of a metaheuristic algorithm to solve a static Dial-a-Ride problem (DARP) with homogeneous requests and a single depot. Indeed, this paper presents the implementation of an Adaptive Large Neighbourhood Search (ALNS) algorithm with some destroys and repair operators tailored for the targeted problem to reach a high-quality solution in an acceptable time consistent with the need of a real healthcare department. This metaheuristic algorithm is compared with the current benchmark in literature and, also, it reports different Key Performance Indicators about the logistic service such as the average time worked and the total waiting time.