«Previous    Next»
Smart Cultivator Sees Weeds, Eliminates Them
A mechanical cultivator that’s as effective as the human eye at identifying weeds in productive crops, then selectively removing them without harming the vegetable seedlings, is being produced by Stout Technologies. The Smart Cultivator uses ultra-sophisticated cameras driven by AI (Artificial Intelligence) to capture high-resolution plant images as it moves through a field, then it instructs mechanically activated blades to remove the harmful weeds.
The Smart Cultivator evolved because a large grower/shipper in the Salinas Valley had several thousand acres of vegetables with a huge weed problem and couldn’t get the labor to work the fields,” says Stout CEO Brent Shedd. “They tried automated weeding machines, which didn’t work because the crop cameras used blob technology that couldn’t differentiate weeds from good plants and did more damage than good. That led them to essentially ‘build their own version of a better mousetrap,’ a prototype of what we’re now producing.”
The Smart Cultivator uses state-of-the-art AI technology that Shedd describes in very simple terms. “Basically, we put the hoe that manual labor uses to weed the fields on a robot and gave it AI, so it knows the difference between weeds and the good crop. The machine is built like a tank, engineered to run hard because it gets kicked around in all types of terrain. It can even be pressure washed without damaging any of the high-tech components.”
The Smart Cultivator mounts on the 3-pt. hitch of a 100-hp. tractor, is coupled to the pto, and can effectively clean and cultivate 1 to 2 acres an hour depending on soil conditions. “The key to our machine is an AI system that identifies weeds with 99.9 percent accuracy, and then mechanical blades eliminate them,” Shedd says. “It doesn’t harm the growing crop.”
Better yet, the Smart Cultivator doesn’t need water breaks, doesn’t get a backache, and can work 24 hours a day because it carries a sophisticated LED lighting system. Strobe lights beside the machine’s vision cameras are timed to fire at the same frame rate at which the machine snaps images, giving the camera the same level of illumination at all levels of light, including darkness. The machine’s modular design allows configurations for vegetable crops such as romaine, leaf lettuce, cauliflower, broccoli, kale, chard, celery, melons, and tomatoes. Single and double-bed models are being produced now, and a larger triple-bed model is being developed. “We sold our first machine in 2020 and now have them working in the U.S., Mexico, Europe and other countries,” Shedd says. Case New Holland recently acquired a minority interest in the company and now provides sales and service to producers around the globe through New Holland dealerships.
“The engineers and technicians who built the machine have diverse backgrounds, with experience in automotive racing, aerospace, defense systems, medical equipment, and agricultural equipment,” Shedd says. “They’ve worked on space telescopes, virtual reality headsets, automated farm equipment, and specialty manufacturing lines. Skills from all these very technical areas were needed to develop the machine.”
The developers beat up the prototype machine over many acres, Shedd explained, then “fixed the weak points, and the machines we’re producing now seldom have issues. The tractor pto charges an alternator to run the electronics. The hydraulics are all self-enclosed, so there are no hoses to hook up.”
Pricing for the machines varies depending on how they’re configured for specific crops. Shedd says cost hasn’t been an issue for selling a machine because once growers see how efficient it is, they realize it can easily replace the equivalent amount or more money that’s being spent on manual labor.
Contact: FARM SHOW Followup, Stout Industrial Technology, 90 Monterey Salinas Hwy., Salinas, Calif. 93908 (www.stout.ai)

  Click here to download page story appeared in.

  Click here to read entire issue

To read the rest of this story, download this issue below or click here to register with your account number.
Order the Issue Containing This Story
2023 - Volume #47, Issue #4