Robots with hoes show farmers possible future of weed control
Go Deeper.
Create an account or log in to save stories.
Like this?
Thanks for liking this story! We have added it to a list of your favorite stories.
Quick Read
In a sugar beet field a few miles east of Moorhead, small four-wheeled robots are rolling up and down the rows of beets.
Powered by a solar panel, the robots use cameras to spot weeds and then guide sharp-edged hoes to scoop the weeds out of the ground.
This is the first year Redmond, Wash.-based Aigen has deployed the robot weeders.
“What we're doing is using a method that's been used for thousands of years” said Aigen field operations director Chris Benner. “Just a hoe to rip them [weeds] out of the ground, with a little bit of extra technology baked into it.”
Turn Up Your Support
MPR News helps you turn down the noise and build shared understanding. Turn up your support for this public resource and keep trusted journalism accessible to all.
Today, farmers rely on herbicides to keep weeds from crowding out the crops they grow. But nature is very adaptive, and weeds are increasingly resistant to pesticides.
Kochia and waterhemp are two common weeds with widespread herbicide resistance.
A single plant that survives can spread thousands of seeds to grow more resistant weeds.
The weed-killing robots are designed to be simple. The goal is to turn them loose and forget them.
However, during this first year of deployment, teams of operators and engineers are working out of trailers in the fields, monitoring the robots and tweaking software and hardware to improve performance.
The robo-weeders look simple as they roll through a sugar beet field at an easy walking pace. There’s a solar panel, four wheels, cameras and two sharp blades on poles.
What's not visible is the artificial intelligence that makes the machine work.
“Trying to understand what a sugar beet is, what a weed is and then pairing that with the ability to strike it while it's moving, that tie in between software and hardware is a very difficult thing to do,” said Benner as he watched a robot work. “It's definitely the bread and butter of what makes the system work."
Aigen co-founder and Chief Technology Officer Rich Wurden is one of the people working in a technology-filled trailer on the edge of a farm field. The former Tesla engineer started this company after a relative who farms in the Red River Valley told him about the problem of herbicide resistance.
Reducing pesticide use is a primary goal for the company.
“There's just not a lot of great alternatives to chemicals,” he said. “Right now we're seeing herbicide resistance and herbicide immunity. And there's no resistance to a steel tool.”
Aigen brought 50 robots to the Red River Valley this summer. They are working only in sugar beet fields. It takes about 10 robots to patrol a 160 acre field. They can weed the entire field in a week working sunrise to sundown. Then they start over.
The company has been keeping a low profile, but Wurden said there’s a waiting list of farmers who want to try robots.
“We've had to say no to way more farmers than we've said yes to,” he said. “We've been doing very little advertising, because the interest in dealing with herbicide resistant weeds is so large.”
The goal is to expand to 500 robots next year, and also start running the machines in soybean fields. Ultimately the company wants to construct a manufacturing plant in the Midwest so the machines can be built closer to a primary market.
Neil Rockstad is a farmer from Ada, and president of the American Sugarbeet Growers Association. The robots are working in one of his fields and he likes what he sees.
“I think they're doing a great job,” he said. “I guess we'll know the results in a few weeks as to how well it controlled the weeds, but my initial response is everything looks really nice.”
Rockstad can recall the days when migrant farm laborers with hoes controlled weeds in sugar beet fields. Then sugar beets were genetically modified to be immune to the herbicide Roundup. That brought an end to hand labor, and weed control was simplified — until the weeds evolved to resist Roundup and other similar herbicides.
Rockstad will evaluate the effectiveness of the robots and the cost per acre before committing to a robotic future. But he has high hopes he can cut herbicide use in the future.
“I would love to live in a world where robotic weeders took over everything and I didn't have to spray a single acre, and I think the consumer would would welcome that,” said Rockstad. “I don't know if we can get there but I certainly don't rule anything out when it comes to technology and our future.”
That’s not likely to happen in the near future. The robots are still learning. Wet weather this year has taught engineers some tough lessons about sticky Red River Valley clay.
But Wurden sees the systems get better every day, and the company is focused on working out the bugs before expanding. The idea is to have a truly autonomous machine that will run by itself and send a message if it breaks down.
“We want to make sure that the farmers really love what we're doing before we start to scale it up,” Wurden said.
“It's pretty impressive how quickly they can take that picture, process that information and then make a strike on a weed,” said Moorhead farmer Trent Eidem.
Eidem has concerns about the cost of buying or leasing the robots, but he's eager to see how the company develops the machines.
“I think agriculture years ago was slow to adapt technology and that's not the case anymore,” said Eidem. “Agriculture is a very tech friendly business to be in and everybody wants to take advantage what can they do to make their season successful.”
In addition to the plan to move the robots into soybean fields next summer, the company is developing other tools it hopes will add value to the machine. The robots could be used for testing soil nutrients, or scanning plants for early signs of disease.
And in the off season, Benner said, solar panels can produce electricity to help power the farm shop.
Want more Fargo-Moorhead news? Check out our local page.