Authors:

Joseph F Norwood

Date Created:

Not specified

Course Title:

Professor:

Not specified

About Paper:

Managing hundreds of acres of farmland often necessitates the use of generalized practices, resulting in inefficiencies in water and nitrogen use. Precision agriculture aims to address this waste through accurate sensing and localized management strategies. Previous research introduced a new configuration of Purdue's agricultural robot, P-AgBot, capable of deploying and reading biodegradable soil sensors; however, this system lacks robust autonomy, requiring significant human intervention. This paper presents advancements to P-AgBot that improve its autonomy, achieving a 90% success rate for sensor deployment and reading, with operation times under two minutes per sensor. Improvements span the mechanical systems such as sensor mounts and chutes; the electrical systems from enclosures to component upgrades; and the software, leveraging deep learning for computer vision. Biweekly field deployments over a two-month period guided iterative refinements and validated system robustness. Results demonstrate that P-AgBot meets the desired performance targets and establishes a foundation for scalable, autonomous soil monitoring. Future work should explore multi- robot swarm deployment to extend coverage and efficiency. Keywords: Autonomous Robots; Precision Agriculture; Sensor Deployment; Computer Vision; Deep Learning

Source:

Purdue University / 2025

Topics:

No topics listed

Co-authors:

Joseph F Norwood