Variable Neighborhood Search Procedures for the Multi-Period Technician Routing and Scheduling Problem

In this paper, we address a new variant of the Multi-Period Technician Routing and Scheduling Problem. This problem is motivated by a real-life industrial application in Telecommunication company. It is defined by a set of technicians having distinct skills that could perform a set of geographically scattered tasks over a multi-period horizon. Each task is subject to time constraints and must be done at most once over the horizon by one compatible technician. The objective is to minimize the total working time (composed by routing time, service time and waiting time), the total cost engendered by the ejected tasks, and the total delay. We propose two variants of variable neighborhood descent, and three variants of Variable Neighborhood Search to solve this problem. Computational experiments are conducted on benchmark instances from the literature. An analysis of the performance of the proposed local search procedures is given. The result shows that our methods outperform the results of a memetic method published in the literature.