Developing the Next Generation of Farmer-Approved Ag Robots

Ask farmers about their hopes and expectations for the next generation of agricultural robotics, and you’re likely to get a slew of responses that are as varied as the farmers themselves. It’s hardly a secret that the ag industry looks a little different everywhere. From large corporation-style plantations that produce dozens of crops to small family-owned viticulture operations, farmers’ needs, desires and challenges depend on a long list of variables.

While machinery manufacturers and technology startups do their best to conduct market research and listen to their customers’ concerns, it’s often difficult to get a full picture of what farmers really want. At World FIRA 2021, the “Farmers, draw me a robot” roundtable sought to bridge this divide by giving four ag professionals from different parts of the world the opportunity to share their unfiltered perspectives.

David Frabotta (Meister Media Worldwide, USA) moderated the panel. He was joined by Chuck Baresich (Haggerty Creek Ltd., Canada), Rohan Rainbow (Grain Producers Australia), Frédéric Brochet (VINOLECT, France) and Wan Adlin Wan Mahmood (KULIM, Malaysia). Because each panelist had a different background and area of expertise, the roundtable discussion covered a range of topics related to what farmers are seeking from their autonomous technologies. This abstract summarizes the main points from the hour-long event.

Customer support is a key component of a successful farmer-supplier relationship.

When it comes to building a fruitful relationship between farmers and technology suppliers, 24/7 customer support is incredibly important. Brochet, whose company farms a 100-hectare organic vineyard, has extensive experience working with Naïo Technologies on their vineyard-specific robot for the past three years.

While the farm attempted to use the robot five years ago, it was unsuccessful due to limitations that made it ill-equipped to handle the vineyard terrain. A few years later, the robot worked mainly on the soil. The farm continued to work with Naïo on improving the robot’s capabilities, and for this vintage, Brochet considers the collaboration a great success. He credits the company’s friendly 24/7 customer support with helping the farm get to this point.

“The first year, we had a lot of problems, so we needed a lot of support,” he says. “We could call the supplier 24 hours, and I don't know how many kilometers they would come to help us because we are quite far away from the production center, but we had a lot of assistance.”

“I think it is key to have someone who is able to help you because it's a totally new technology. It's not like a tractor where you can find people who knows about it,” Brochet continues. “For this collaboration to work, you need to have a really good relationship with a supplier, and you need to have a 24-hour support absolutely available with people who are nice and open-minded.”

Farmers can work to accommodate robots, but they can’t change everything.

One thing the panelists agreed on about working with robotics suppliers is that farmers are willing to make changes that will help autonomous machines adapt to their operations, but suppliers need to be realistic about what can be done and how long it will take to complete.

For example, Mahmood, who has a lot of experience working on plantations, mentioned farmers who have been operating the same way for the last 50 to 60 years, aren’t able to make wide-scale changes to their planting systems or processes overnight. Instead, Mahmood says that integrating streamlined technologies that solve specific pain points are more likely to be more successful than a machine that requires a big transition and doesn’t address the farmers’ biggest concerns. He recommends keeping the technologies simple.

Mahmood does believe, however, that there are also small adjustments farmers can make to integrate robots into their current workflows. Fertilizer applications are one way to use robots effectively. Instead of broadcasting the application, robots can help deliver specific amounts of fertilizer to each plant—without requiring the same level of manual labor. Empowering farmers to cut costs on labor and inputs is incredibly valuable, Mahmood says.

“Return on investment is in the eye of the beholder.”

This insight comes straight from Baresich, who farms 3,500 acres of cash crops in Southern Ontario with his brother and wife. Though farmers tend to be cash conscious, Baresich believes there are many variables that go into determining whether a robot is worth the investment. Manufacturers, startups and suppliers that want to appeal to farmers need to be aware of what he calls “the intangibles”—especially on single-use robots.

“It’s not just about the money because there are other intangibles that are hard to value,” Baresich says, noting that the ROI for autosteer on a tractor is basically nonexistent but the technology is ubiquitous. “It comes down to what can be added to a single-use robot that only does one thing. Then, I’m measuring what solving that pain point is worth. What's the cost of the current treatment that we're doing? What is the cost of my new solution as well as maintenance and upgrade costs? And perhaps the biggest thing I worry about with these machines is the obsolescence factor.”

Baresich compares the robots to the cellphone: People prefer the new model, and they aren’t about to buy the old one right before the new version comes out. Ultimately, Baresich says, he considers the investment worth making if the cost per acre using the robot is the same but enables him to get a greater bang for his buck from his workers.

“As an example, this year, we used the OMNiPOWER in conjunction with an existing Terregator,” he says. “So, the person operating the Terregator was operating the OMNiPOWER at the exact same time. Now, we’re getting twice as much done with the same staff. That’s where our focus is right now, on optimizing the value of our labor.”

Connectivity and functionality are farmers’ top pain points.

As farmers begin integrating robots and autonomous machines into their operations, there are still obstacles that need to be overcome. Every producer will need to be selective about the size, weight and cost of the machines that they use, given the size and scale of their farms. These considerations do not address some of the bigger pain points operators will encounter, such as connectivity issues, training gaps and whether the machine will function as promised.

“Probably more than half the farmers in Australia have no access to cellular phone connectivity,” says Rainbow, who works closely with the Australian grains industry. “That's actually quite a challenge if you want to service your machine or just run diagnostics on whether this machine is performing correctly and providing that information back to the operator. You can get around some of these things from a safety point of view by using HF radios and those sort of things, but it doesn't provide you the granularity of information or the richness of data that some of these robots are using. To really get the value of the robots, we want to collect information and data from the field. Getting that back to the user is probably one of the key challenges.”

Connectivity isn’t the only thing that can limit a robot’s functionality. There’s also the complexity of the task at hand. While Baresich has found great success using machines for simple repeatable tasks, he’s run into issues with technologies that require more advanced algorithms.

“There's some oddities built into the robots that create challenges,” he says. “For example, the robot’s mission-planning software will send a send a robot on a mission that it can't complete, such as actually navigating a specific turn in the field. Sometimes, we know as a user with a joystick that the robot can navigate that turn, but the software doesn't allow due to some kind of safety protocol that might be built. So, some of that work has been frustrating.”

Baresich encourages manufacturers to test their robots based on real-world scenarios—ones that are often far more challenging than those simulated in a lab environment.

“Sometimes I think that the testing that gets done on the robots is limited because the manufacturers are so invested into the robot themselves that they designed their tests around scenarios that they know the robot can complete,” he says. “So, the testing might be 95% successful in those scenarios, but I want to put the robot into a typical farmer scenario and tell it what I want it to do. That's when we uncover some of these logistical and mission-planning challenges, where the engineers will say, ‘Well, why would you do that?’ And that’s where I say that I shouldn't always have to modify what I want to do to fit the robot. There needs to be communication between the engineers and the farmers to close this gap.”

Another challenge is the divide between what farmers know and what robots can be taught. The panelists reflected on the generational intelligence that farmers gain over many years of growing the same crops. This information—details about everything from soil moisture to how fast a tractor can safely move through the rows—can be difficult to program into a machine. Baresich recommends that farmers begin documenting their knowledge before bringing autonomous technologies onto their operations. This information can then be integrated into the robot’s programming ahead of time.

On the engineering and product development side, farmers want to see the technologies deliver on their promises. Baresich gives the example of sensor technologies. While there are a lot of sensor capabilities, some have yet to meet the expectations set by the manufacturers.

“There's lots of issues with LIDAR sensors that basically become ineffective in the sun or in certain conditions, and they just aren’t living up to the promises being made,” he says. “And it's not that they won't, it just that we're not quite there yet.”

Baresich says the other consideration manufacturers need to address is how their technologies work with others in the field. Setting industry standards will be important for farmers, who often work with a wide range of equipment from a variety of manufacturers.

“I would like to see some degree of agreement between the manufacturers on interchangeability of sensors, like a plug-and-play platform, where if I need a certain sensor on my robot, that's a little bit different than what it came with, I should be able to swap that out to get a different band or camera or these types of things. Unfortunately, I see a lot of manufacturers developing their own sort of private proprietary things, which is great for them. But if they only ever build five of them, what good does it do? The rest of us that may never purchase that robot.”

Farmers and suppliers can work together to change perceptions about autonomous ag technologies.

The limitations of today’s ag robotics have not dampened farmers’ hopes for what’s to come. They foresee a future where autonomous machines are commonplace on many operations and are finding new ways to bring their customers and neighbors into the fold. Brochet sees this effort as a helpful way to help rebuild people’s connection with the land.

“When people see big tractors, they are afraid of them, or they see big sprayers and think we are polluting,” he says, noting that as organic farmers, this perception is not entirely accurate. “So, we’ve been willing to use this robot to connect people from the village back to the land and show them how farming might be done in the future. They get excited and want to talk to the wine maker and the wine grower. People don’t do this when you’re on a tractor because the cab closes you off, but when you work in the field alongside the robot, people come and talk to you. It is valuable for this project to have a lot of positive images and communication. We expect a lot of good things on this side.”