Aarhus Universitets segl

Operational, Economic, and Environmental Assessment of an Agricultural Robot

Publikation: Bog/antologi/afhandling/rapportPh.d.-afhandling

The use of robotic applications in agricultural machinery systems has grown significantly in recent years. This is aimed at improving efficiency, saving time, conserving energy, and increasing crop yield through advanced precision agriculture methods. Progress in mechanical engineering, sensory perception, computing, and human interfacing has contributed to this advancement. To meet future agricultural demands, research, and development are essential, as emphasized by the Danish National Robotic Strategy (2020). The strategy calls for integrating different disciplines and designing robot systems specifically for digital farms. However, a pressing issue is the lack of sustainability evaluations that consider economic, operational, and environmental aspects when implementing autonomous systems. It is crucial to assess the complex interactions between machinery, biological material, and the dynamic environment to accurately measure the performance of robotic systems. Bridging this gap requires interdisciplinary expertise, combining knowledge from the bio-production domain (agriculture, horticulture) with computer science and research in operations and management sciences (modeling, optimization, and simulation).
This Ph.D. thesis focuses on assessing an agricultural robot from an economical, operational, and environmental point of view. To this end, several in-field operations of this robot including seeding and weeding were monitored, measured, and analyzed and the results were compared with the conventional machinery. Moreover, a 3D field area coverage planning model was developed to optimize the robot's in-field operation in different scenarios (seeding, weeding), estimate the agronomical factors such as crop health (based on the NDVI index), and evaluate the environmental impact such as the risk of soil compaction in a defined field sample for the studied robot. One of the important operational factors is the amount of fuel consumption. Machine learning techniques were applied to develop a model for estimating the fuel consumption of this robot based on the defined size of the field. Furthermore, a simulation model was developed in order to simulate the robot's operation in different scenarios (seeding, weeding) to estimate performance (task time, capacity, energy) defined for specific field sizes for one robot or fleet of robots.
ForlagAarhus University
Antal sider229
StatusUdgivet - 2023

Se relationer på Aarhus Universitet Citationsformater

ID: 337545267