Automated methods have the potential to enable sustainable agroecosystem development by reducing labor requirements. We propose planning algorithms for robotic sampling to enable resource accounting and precision agriculture operations. Our system utilizes geostatistical principles to model dynamic processes in agriculture. The algorithms prioritize sampling in areas with higher uncertainty, leveraging previous information and field knowledge. We plan field trials and system improvements, allowing the use of observations from previous surveys and multiple autonomous agents. Our work will be released under a free and open license, encouraging student engagement and experimentation in managing lands using these methods.
Automation, Cyberphysical Systems, Robotics , and Mechatronics
