Variable Neighborhood Search based algorithms to solve a Multi-Attribute-k-Travelling Repairmen Problem

The technician routing and scheduling problem (TRSP) is a special case of the k-traveling repairmen problem (k-TRP), where each technician has a specific skill. The k-TRP is an extension of the vehicle routing problem with time windows [2]. Despite their numerous practical applications, k-TRP [1] has received limited attention in the literature compared to the TRSP [3]. First, we studied the MA-TRSP already presented in [6] encountered by a company providing maintenance and repair of electronic transaction equipment. Technicians have different skills levels which allow them to perform different tasks. Along the route, a pre-assigned depot can be used (one per day) to replenish the inventory of spare parts and also get, if needed, special parts. Each task has a service time and multiple possible time windows to be served. It has also a gain which stands for its priority. Given that, not all tasks can be served. The objective is to maximize the total gain, minus the total traveled distances and the total overtime over all routes. Second, we studied the same problem where technicians have multi skills, which becomes more difficult to solve due to the increased flexibility in the assignment of technicians to requests. The new MA-k-TRP problem considers polyvalent repairmen.