Researchers from the University of Idaho Coeur d’Alene have been granted a $139,000 award from the U.S. Department of Agriculture (USDA). These funds will fuel the development of a robotic weeding system at the U.S. Forest Service’s tree seedling nursery in Coeur d’Alene, which aims to provide quality seedlings for publicly owned lands and develop the best possible methods for producing quality seedlings.
This project, known as Project Evergreen, was inspired by a desire to investigate opportunities for enhancing nursery operations through automation.
The project will be led by Garrett Wells, a UI doctoral student in computer science. Contributions from John Shovic, director of UI’s Center for Intelligent Industrial Robotics and associate research professor, and Mary Everett, a postdoctoral researcher for the center, will help the project’s development as well.
The robot, named Amiga, will be produced by a company called Farm-ng and retails for around $13,000.
The robot will take on expensive, labor-intensive tasks so that more seedlings can be grown at nurseries, according to Wells. Project Evergreen is funded through the Repairing Existing Public Land by Adding Necessary Trees (REPLANT) Act specifically, which focuses heavily on reforestation. Project Evergreen will be used to investigate ways to improve nursery operations and increase the number of seedlings raised for reforestation.
One way to improve the current state of nurseries is to focus on weed control and ways to lower the stress put on growing seedlings. The most popular ways of weeding presently are the use of herbicides and hand-weeding. However, as explained by the USDA, herbicide programs vary from different tree species, and they also require an application that avoids seedling injury while still targeting weeds. Hand-weeding is usually a strenuous and time-consuming job. Project Evergreen aims to find a solution that can avoid both.
“This robot will hopefully demonstrate an effective non-chemical alternative for weed control,” said Wells. “Treating weeds with the robot before they mature will help reduce weed pressure over time by eliminating latent seeds (seeds that do not sprout instantly) which germinate throughout the growing season.”
The robot is also superior to other forms of weed eradication when it comes to the weather. Hand-weeding can become entirely impossible in certain weather conditions, and the spraying of herbicide can drift and can be impeded in the presence of wind. While the robot’s effectiveness may be reduced by precipitation that causes muddy field conditions, it can still accomplish its job no matter how cold or windy it is.
As far as looks go, the robot has four wheels with large treads, a square frame, a seat on top of the frame, and a touch screen at the front. It will come with an embedded computer that “makes geo-tagged data collection efficient and precise,” as stated by farm-ng.com. The data collected by the computer will be accessible through a custom app, and visualization of said data will be appropriately displayed.
“This project is a great example of how robotics and AI can play together to make magic,” Shovic told UI in a news release. “The U.S. Forest Service is looking to dramatically increase their seedling production in the future and this robot will be a great help. The University of Idaho is proud to be part of this path.”
Rebekah Weaver can be reached at [email protected].