Implementation and applications of harvest fleet route planning

In order to support the growing global population, it is necessary to increase food production efficiency and at the same time reduce its negative environmental impacts. This can be achieved by integrating diverse strategies from different scientific disciplines. As agriculture is becoming more data-driven by the use of technologies such as the Internet of Things, the efficiency in agricultural operations can be optimised in a sustainable manner. Some field operations, such as harvesting, are more complex and have higher potential for improvement than others, as they involve multiple and diverse vehicles with capacity constraints that require coordination. This can be achieved by optimised route planning, which is a combinatorial optimisation problem. Several studies have proposed different approaches to solve the problem. However, these studies have mainly a theoretical computer science perspective and lack the system perspective that covers the practical implementation and applications of optimised route planning in all field operations, being harvesting an important example to focus on. This requires an interdisciplinary approach, which is the aim of this Ph.D. project. The research of this Ph.D. study examined how Internet of Things technologies are applied in arable farming in general, and in particular in optimised route planning. The technology perspective of the reviewing process provided the necessary knowledge to address the physical implementation of a harvest fleet route planning tool that aims to minimise the total harvest time. From the environmental point of view, the risk of soil compaction resulting from vehicle traffic during harvest operations was assessed by comparing recorded vehicle data with the optimised solution of the harvest fleet route planning system. The results showed a reduction in traffic, which demonstrates that these optimisation tools can be part of the soil compaction mitigation strategy of a farm. And from the economic perspective, the optimised route planner of an autonomous field robot was employed to evaluate the economic consequences of altering the route in selective harvesting. The results presented different scenarios where selective harvest was not economically profitable. The results also identified some cases where selective harvest has the potential to become profitable depending on grain price differences and operational costs. In conclusion, these different perspectives to harvest fleet route planning showed the necessity of assessing future implementation and potential applications through interdisciplinarity.