Intervention schedule optimization with travel time minimization for a Value-Added Reseller by solving the Capacitated Vehicle Routing Problem

Abstract

With the significant increase of service providing companies and the option of in home installation or maintenance, the importance of finding the optimal planning for the workers has risen accordingly. Global warming, high fuel prices, and important labor costs call for the need to minimize travel and working time and reduce the impact on the environment. In this paper, the CVRP is solved to establish a planning of interventions, being installation and maintenance, at customers of a value-added reseller (VAR). The goal is to minimize total travel time, maximize labor time per day, combine jobs that need two workers in the same van, and to reduce emissions. In contrast to previous research on routing optimization, limits are set to both the working time and the sum of the working time plus the travel time. In addition, it centralizes installations that need two workers on the same route, further minimizing the use of vans. As a result, scheduling becomes faster, more accurate, and scalable, leading to a significant reduction in overall asset and labor cost, and to less $C{O}_2$ emission, thus cleaner logistics. This intervention planning is compared with the random planning and planning proposed by an expert planner. Applying our algorithm on various configurations of 16 to 82 customers led – in a time span of seconds – to a relative gain of 3% for the smallest application and up to 38.6% for the largest one, compared to the time-consuming planning made by the expert human planner. Moreover, to visit 82 customers 3 less vehicles are needed (21 instead of 24), in comparison to the human made schedule.