Andrew Landers

A fruit tree sprayer fitted with sonic sensors to determine the absence or presence of trees and their height. Cornell researchers are working on developing a completely robotic sprayer that wouldn't require a human operator, thereby preventing exposure to pesticides.

In coming years, the bucolic scene of a driver guiding a tractor and spray rig up and down an orchard or vineyard could go by the wayside. Researchers at Cornell's New York State Agricultural Experiment Station (NYSAES) are helping to create robotic tractors and sprayers that do not require human operators.

Andrew Landers, a pesticide application engineer at NYSAES, is part of a $3.9 million U.S. Department of Agriculture-funded project to develop, test and evaluate a fleet of autonomous tractors designed for precision agriculture applications. The project, Integrated Automation for Sustainable Specialty Crop Farming, is in collaboration with the National Robotics Engineering Center (NREC) at Carnegie Mellon University.

Landers leads a team of Cornell researchers in developing an automatic sprayer that recognizes a target -- in this case canopy volume -- and adjusts air and liquid deposition of chemicals accordingly. NREC will focus on developing the tractor.

"I will be providing details on how to monitor the spraying operation and which factors will improve sprayer output," Landers said. "We will then design the automatic sprayers to be fitted to the autonomous tractors. One of the major challenges, then, will be getting the tractor and the sprayer to talk to one another."

Robotics has revolutionized many fields -- space expeditions (the Mars Rovers), deep-sea submersibles and pilotless drone aircraft are some prime examples. Like these vehicles, which were designed to operate in and around hostile or dangerous environments, the autonomous tractor-sprayer combination removes the driver from the danger of exposure to toxic chemicals.

Goals for the project include developing tree-level precision agriculture applications that leverage, at very low cost, autonomous mobile platforms and supporting infrastructure; reducing the cost for wide-scale adoption; and soliciting feedback from growers, regulators and technology suppliers. The researchers will also study such questions as how disease detection, yield estimation and precision spraying can be most effectively deployed from the mobile platform; how many platforms one operator can safely monitor and what the installation, setup and support issues are associated with the system.

John Deere will deliver four tractors for testing at Southern Gardens Citrus in Florida. Brad Muise, a postdoctoral associate in Landers' program, will evaluate the physical properties of deposition and reduction of drift annually. Although this phase of the work will be done in orange groves, the technology will be easily transferred to orchards and vineyards. Researchers hope to deliver a prototype in one year and complete the project in three years.

Joe Ogrodnick is a writer at the NYSAES in Geneva, N.Y.