Improving the manual harvesting operation efficiency by coordinating a fleet of N-trailer vehicles

In the last few years, automatic guidance of agricultural vehicles has received increased interest to improve field efficiency, while releasing human workers from monotonous operations. In this context, this work proposes a short or medium-term alternative to complete automation of manual harvesting processes by introducing a fleet of robotic N-trailer vehicles to support the crop transportation task. With the aim of coordinating several vehicles which share the workspace with human pickers, this work presents a decentralized cooperative navigation strategy that divides the field into harvesting areas, allocates an active N-trailer vehicle per area, determines the number of trailers to use according to the payload capacity and maneuverability constraints, and coordinates the departure time of backup N-trailers in order to reduce non-productive times. The proposed strategy includes a harvesting sequence generation stage based on centralized global information, and a dynamic route planning stage based on both global and local information. In order to make the navigation strategy robust against the variability on the harvesting rate and the uncertainties about the field productivity, the global information is continuously updated based on local information from the vehicles. A simulator was built in order to evaluate the performance of the proposed strategy in realistic scenarios having different field productivity and machinery availability. Thus, by using a large fleet of N-trailers and properly coordinating them to share the workspace, the results showed an efficiency improvement of up to 82% with respect to the basic case where two vehicles (one active vehicle and one backup vehicle) were used to cover the whole field, as reported in the literature