Stacey Santos
The robots have arrived to help with weed management on the farm, but where do you start? How do you know what’s right for your crop, your acreage–and your budget?
Thankfully, a number of options have emerged for managing agricultural tasks with robots, along with a crop of experts who are working on finding Canadians the most promising options for robotic weed control.
In a recent episode of the Organic BC Podcast, host Jordan Marr sat down with two members of Ontario’s AgRobotics Working Group, who explained why robots have become a viable option for farmers, including organic producers in BC who are looking to boost sustainability, soil health, and efficiency on their farms.
AgRobotics Working Group
Since 2021, an AgRobotics Working Group, led by Haggerty AgRobotics and the Ontario Ministry of Agriculture, Food and Rural Affairs (OMAFRA), has paved the way for the use of autonomous robots in farming. The group, which also includes growers, grower associations, agri-business, universities and colleges, federal and municipal governments, and technology companies, serves multiple functions, but foundationally it exists to help farmers make smart decisions as they consider investments in tools that can be really expensive.
“We’re trying to basically ground truth so producers do not have to be guinea pigs,” said Kristen Obeid, weed management specialist with OMAFRA and chair of the AgRobotics Working Group. “We want to work out all the bugs first, because this is complicated.”
“We want to deal with challenges that the farmers want us to solve,” added Chuck Baresich, farmer and president of Haggerty AgRobotics. “We look through our network and say, ‘Does anyone have an idea of a solution that may or may not solve that problem? Let’s go test that and see.’”
Why robots, and why now?
Part of the answer is that the technology has developed to the point that some of these tools are just plain good at delivering on what they promised to do, and meanwhile they’re starting to be able to compete on cost with other methods of weed control.
But, there has also been a major development in agriculture that motivated farmers to give these tools serious consideration: herbicide resistance.
“Herbicide resistance is going to be the biggest challenge facing agriculture in the next decade,” said Kristen. “In Ontario alone, we have 23 resistant weed species.”
From a chemical perspective, weed control options are not as available or as effective as what they used to be.
“Resistance is on a trajectory,” Kristen added. “It’s not going away. We need to find different integrated weed management solutions for growers to implement and not continually rely on herbicides.”
But what about good old fashioned mechanical cultivators pulled behind tractors driven by farmers, aided by RTK GPS? Why should farmers consider robots instead of those tools?
Soil damage.
A lot of these robots incorporate artificial intelligence, so they only take out the weeds, with minimal soil disturbance. And, the robots are often lighter than traditional tools, so there’s a lot less soil compaction.
And then, there’s the potential for these robots to save farmers the precious commodity of time when compared to traditional mechanical cultivation.
“That’s where the allure of robotics comes in,” said Chuck. “RTK and auto-steer systems have got pretty good reliability and trustworthiness, and we’re at the state now where we can take that knowledge and move it into an unattended machine.”
Let’s talk about cost
The costs are variable. Kristen and Chuck discussed six robots, and the price varied from $60,000 to a quarter million (CDN), depending on the robot.
What matters most, anyway, is cost effectiveness and value proposition–also difficult to pinpoint because the value of one of these machines really depends on the context of any given farm.
“These robots will never compete against a 120-foot-wide sprayer,” said Chuck. “Where the robots compete is if, say, you’re used to bringing in a manual weeding crew. Or traditionally, when a farmer has to cultivate, they take either themselves or one of their top workers and put them in the tractor. They can be doing something else of more value than sitting in a tractor seat, whether that’s seeding a different crop or monitoring for disease or something else.”
“This robot has nothing else to do, so it actually doesn’t matter that it’s moving slowly, as long as it gets over the field.”
Tools under evaluation
The FarmDroid
One of the tools the AgRobotics Working Group has been evaluating is the FarmDroid, a 900 kg, solar-powered robot that uses a geolocation seeding mechanism to plant in a perfect grid pattern across your farm. Knowing where every seed is, the FarmDroid then weeds that exact same field after planting, before the seeds even come out of the ground.
It’s completely autonomous–it charges itself, runs through the night, and doesn’t require any input from the operator other than if there’s a problem.
The tradeoff for its precision is that the FarmDroid is very slow–its top speed is 0.8 km per hour and it can only handle about eight acres every 24-hour period. If you’re a larger producer and wanted 50 acres of a crop planted on the same day, you’d need multiple FarmDroids running at the same time.
You also need to be cognizant from an operational perspective that you also only have eight acres a day for the weeding process. You need to make sure the field you’re putting it in is appropriately sized so the FarmDroid can get to all the weeds before they grow too large to be taken out.
AgroIntelli Robotti and Naio Orio
The FarmDroid is best suited to higher value horticultural crops like carrots or beets, but is too low to the ground, too small, and too slow to make much economic sense for field crops like soy and corn.
For those kinds of crops, the AgRobotics Working Group has been evaluating different robots, including the AgroIntelli Robotti and the Naio Orio.
These machines are quite similar to each other and can go through crops that are waist high or more, and can do four to six 30-inch rows at a time.
The only difference between the two is that the Naio Orio is an electric robot that can get eight to ten hours per battery charge, and that the Robotti has a little diesel engine that can run for about 60 hours on a tank of fuel.
Both are really precise machines with camera guidance and three-point hitch attachments, so they’ll fit your existing cultivator or other implement. They can also accommodate larger acreage: the Orio can keep a 50-acre field under control, while the Robotti is closer to 100 acres.
Nexus La Chèvre
Jumping back to horticultural crops, another interesting tool is La Chèvre (“The Goat”) from Nexus.
La Chèvre has a series of cameras at the front of the machine that are programmed to know what the crop is. If it sees anything that’s not lettuce, for example, it sends a pair of mechanical grippers down to remove whatever the obstacle is.
The real advantage is that this robot can be trained to leave microscopic weeds alone, resulting in less soil disturbance. And, it also has the advantage of being able to pull out slightly larger weeds, which the other systems cannot do.
“Where we’re seeing some interest in a robot like this is from farmers who might set up a herbicide program where they put down some form of residual and then use the robot from that point on to do weed control,” said Chuck. “Generally speaking, they’re trying to get the robot up to five acres in a 24-hour day–approximately three workers worth.”
Options for no till systems
For the most part, the robots that Kristen and Chuck deal with aren’t a great fit for a no-till system, and one of the main reasons is what Chuck calls “trash”—the residual roots, stock and leftovers from the previous crop that plug up the tooling and are too much for the cultivator to deal with.
The AgRobotics Working Group is working on solutions: they have one grower doing a trial with the Robotti in a strip-till system, carrying out traditional spraying in the strips and then using the robot closer to the crop.
They’re also looking at tooling options and moving into alternative technologies, whether it’s laser weeders, spot sprayers, or electrical weeders for a more precise weeding system. But, these systems are much more costly and complicated.
“Again, you talk about trade-offs,” said Chuck. “Nothing is cheaper than a cultivator, and the second you start moving to vision systems, AI systems, cameras, mechanical arms, lasers, whatever it is, all of a sudden it gets very costly, very quickly.”
Robots aren’t magic
While Kristen and Chuck are optimistic about the role these machines can and will play on the farm, they do come with limitations.
For one, your soil conditions have to be such that the robot can accommodate it. For example, in wetter conditions, the FarmDroid operates better and covers more ground on sandier soil than clay.
And then there’s the internet. With all of the RTK GPS technology, you need to be reliably connected at all times, which can be a barrier.
Another hurdle is the investment, which Chuck describes as a tough pill to swallow.
“Most people aren’t rolling around in cash,” he said. “You know they’re sitting there saying, ‘Hey, I already spent my $300,000 or $400,000 on a tractor, or on this piece of equipment, or whatever. And you’re asking me to do that again and abandon my initial investment?’”
“There are very legitimate reasons to be hesitant.”
Adding to the fear is the constant improvement of technology (in five years time, there’s going to be an upgraded version) and questions around the service (if I’m buying a robot from France, what happens if it breaks down?).
But there’s good news: Businesses are popping up across the country that are trying to solve these issues and provide service to producers who purchase the equipment.
Success stories
While there are some considerations to keep in mind, both Kristen and Chuck are extremely optimistic about the potential of these tools. Here are some success stories they observed in their trials:
From Kristen: “In the Holland Marsh area of Ontario, we have a fig leaf that is now resistant to two groups of herbicides that growers use in their carrot and onion production rotation. We’ve had tremendous success with the FarmDroid. We’ve tested in sugar beets, and then this past year we trialed it in onions and again, with great results.”
“The other one that we had last year was planting onions with the FarmDroid and getting a more consistent size of onion sets,” she added.
From Chuck: “One of the other robots we’ve used is called a Naio Oz, and we had a great experience with one of our customers using it to mark out rows of garlic, and then plant the garlic and weed it over the summer. That was a great use case of a scenario where the farmer’s using the robot like an assistant.”
A lot of potential, but beware of the hype
When doing your own research into agricultural robotics, don’t get sucked into the hype around AI and machine learning, even when robotics companies make wonderful claims on their websites.
“Not all of the stuff we’re doing works,” said Chuck. “I’m a very optimistic person and I love the technology, but boy, I don’t want to oversell it.”
“I think the market will pick the winners and the losers, but I think the technology is very, very promising,” said Kristen. “There are still lots of questions and lots of hurdles before there will be widespread adoption. But I think we’re on a good path forward.”
To learn more about any of the tools discussed here, or to learn more about the AgRobotics Working Group, visit:
AgRobotics Working Group
Haggerty AgRobotics
FarmDroid
Naio Orio
Nexus La Chèvre
This article was funded through Sustainable Canadian Agricultural Partnership, a federal-provincial-territorial initiative.
Opinions expressed in this article are those of the author and not necessarily those of the Government of Canada or the BC Ministry of Agriculture and Food. The Government of Canada, the BC Ministry of Agriculture and Food, and their directors, agents, employees, or contractors will not be liable for any claims, damages, or losses of any kind whatsoever arising out of the use of, or reliance upon, this information.
Stacey Santos is the Communications Manager for Organic BC. She lives, writes, and gardens in the beautiful and traditional territories of the Lekwungen peoples, who are now known as the Esquimalt and Songhees Nations.
Featured image: The Naïo Orio robotic weeder. Credit: Naïo Technologies.
