Using the Vehicle Routing Problem to reduce field completion times with multiple machines

Abstract

The Vehicle Routing Problem (VRP) is a powerful tool used to express many logistics problems, yet unlike other vehicle routing challenges, agricultural field work consists of machine paths that completely cover a field. In this work, the allocation and ordering of field paths among a number of available machines has been transformed into a VRP that enables optimization of completion time for the entire field. A basic heuristic algorithm (a modified form of the common Clarke-Wright algorithm) and a meta-heuristic algorithm, Tabu Search, were employed for optimization. Both techniques were evaluated through computer simulations in two fields: a hypothetical basic rectangular field and a more complex, real-world field. Field completion times and effective field capacity were calculated for cases when 1, 2, 3, 5, and 10 vehicles were used simultaneously. Although the Tabu Search method required more than two hours to produce its solution on an Intel i7 processor compared to less than one second for the method based on Clarke-Wright, Tabu Search provided better solutions that resulted in reduced field completion times and increased effective field capacity. The benefit provided by Tabu Search was larger in the more complex field and as the number of vehicles increased. With ten vehicles in the real-world field, the benefit provided by Tabu Search over the modified Clarke-Wright resulted in reduced completion time of 32%, but even with only three vehicles a 15% reduction was obtained. While ten vehicles may only be applicable with future autonomous machines, simultaneous usage of three machines is not uncommon in current production. As producers consider using multiple machines to improve field completion times and effective field capacity, optimization of the vehicle routing will play an important role in ensuring those improvements are fully realized.