AI-Powered Sprayer Targets Weeds with Precision Herbicide Use
A new hooded sprayer from Verdant Robotics uses artificial intelligence and machine learning to identify and target individual weeds-redefining herbicide application efficiency in agriculture. With 3D mapping and pinpoint accuracy, it's a glimpse into the future of sustainable crop protection.
A new AI-powered sprayer demonstrated recently in Indiana may mark a turning point in how herbicides are applied on row crops. Developed by Verdant Robotics, the hooded machine uses artificial intelligence, cameras, and machine learning to detect and treat individual weeds-creating a new frontier for precision agriculture.
At Becknology Days in Atlanta, Indiana, the sprayer amazed attendees by targeting simulated weeds-cover crops planted among soybeans-and spraying them with water. If the sprayer had used herbicide, the weeds would have been eliminated with near-surgical precision.
"We have 22 of these machines operating this summer," said Chad Yagow, Verdant's Director of Digital Agronomy. While most units are active in California vegetable operations, a few are now being piloted in the Midwest, including on a commercial organic farm.
A light under the hood of Verdant Robotics' sprayer helps cameras develop precise targets for shooting herbicide at weeds.
The sprayer builds a 3D weed map of each field. Using that map, it outlines a virtual box around every plant. The onboard system then calculates the required herbicide dose and applies it with dedicated nozzles-adjusting for weed size and shape. "If a weed is bigger, it gets more herbicide," Yagow explained.
But the real breakthrough lies in what the machine doesn't spray. The system will not fire through a virtual box, protecting nearby desirable plants like soybeans from unintended application.
Each camera on the unit is paired with four dedicated nozzles, and the current model operates at 3.5 km/h, with a rate of up to 196 spray shots per second per row, covering 4.5 meters per pass.
While not yet approved for use on large-scale corn and soybean fields, Verdant is working to scale up the technology. Yagow noted their next targets include boosting computing power and raising the operating speed to 6.4 km/h, with an ultimate goal of 13 km/h, making the machine viable for broadacre crops.
This perspective shows underneath the Verdant sprayer, looking at the cameras. There are four nozzles dedicated to each camera.
Regulatory alignment is also essential. Current herbicide labeling does not permit some products-like glyphosate-to be used in such targeted ways. However, other herbicides such as glufosinate can legally be applied via this method.
"We follow label guidelines. It's why we can't spray glyphosate yet, but glufosinate is allowed," Yagow said.
The prototype's success demonstrates the growing role of robotics in agriculture, particularly in boosting input efficiency, reducing chemical drift, and protecting crop health.
Cover crop plants sown with these soybeans represent weeds in this demonstration plot. The row on the left was not sprayed. Note the moisture on non-soybean plants in the row on the right. Only water was used in this demo.
As weed resistance, environmental regulations, and rising input costs reshape row crop economics, tools like Verdant's sprayer could offer a game-changing advantage to farmers seeking precision and sustainability.