The Agricultural Routing Planning in Field Logistics

Introduction

The globalisation of the agricultural sector has presented farmers with stronger competition, impacting especially on the market prices of their produce. Advancements in agriculture are generally associated with advanced seeds (Murphy, 2017), use of fertilisers, better resistance against diseases and pests, and water utilisation (Jones et al., 2017). While factors such as these potentially improve the productivity of cultivated areas, they do not address the deployment of humans and machinery for the purposes of harvesting and transporting the produce from the fields to storage areas. Several benefits are derived from optimising the utilisation of humans and machinery, namely, a reduction of time and cost thereby increasing profit. The cost of agricultural machinery is reduced; fewer seasonal workers need to be hired, and less time is required for bringing in the produce. This last factor is critical as the sizes of farms are increasing, yet the time window for harvesting remains the same.

Here, we set the focus on agricultural field logistics. This is related to the optimisation of the vehicle movements necessary to manage the fields including seeding, fertilisation, and harvesting, coordinate multiple vehicles as well as the movement between produce storage, field, and depot. Despite the resemblance with the VRP or TSP, the specific configuration of the problem as well as the utilisation of the vehicles justifies the introduction of a new problem class. Hereafter, we call this kind of agricultural machine planning problem “Agricultural Routing Planning” (ARP). The ARP addresses the movement and route planning of vehicles or machines to undertake several agricultural tasks on the fields. The assumption is that the field is traversed by a series of established tracks (orchards, vineyards, or agricultural fields with symmetrical planted seeds like crops or potatoes) that is accessible to larger agricultural machinery including harvester and possible accompanying agricultural vehicles. The existence of structured tracks supports the human drivers but also allows the operation of auto-steering agricultural machinery using programmable navigational computers and GPS guidance (Barrientos et al., 2011). The routing and planning optimisation is targeting a minimisation of the total distance driven to manage fields, reduce the number of vehicles, and synchronise multiple vehicles. Decisions to make include the selection of vehicles based on capacity and size, selection of areas to operate on, and order of tracks to traverse (Bochtis & Vougioukas, 2008).

Various models and optimisation methods for the ARP can be found in the literature. Multiple authors (Bochtis & Sørensen, 2009; Conesa-Muñoz, Bengochea-Guevara, Andujar, & Ribeiro, 2016; Gracia, Velazquez-Marti, Estornell, Vel Azquez-Martí, & Estornell, 2014; Zhou, Leck Jensen, Sørensen, Busato, & Bothtis, 2014) demonstrate how the agricultural routing planning problem can be described as Travelling Salesman Problem (TSP) or the Vehicle Routing Problem (VRP), which are the most common formulations for routing planning. The TSP has the objective to determine the minimum cost of routes that must be travelled by a salesman(s) from a depot to a group of geographically dispersed customers. The TSP is often found in many real-world problems, including many in the agricultural area. The extended version of TSP is called VRP. VRP is related to plan the routing of vehicles with the same objective as TSP. In agricultural logistics, especially in field logistics, the TSP and VRP often occurs in the planning of the movement of machines to undertake multiple agricultural tasks.