Pictured Above: During Agritechnica in Hannover, Germany, AGCO’s Fendt brand introduced Project Xaver, a swarm of a dozen or so small, autonomous seeding units, replacing one 8-row planter. The small autonomous units reduce compaction and are also designed so that if one goes down, another unit can step in to finish the job.
Ask someone in agriculture what comes to mind when they think of autonomy and they’ll likely say a driverless tractor.
But many experts in precision farming say that simply removing the farmer from the cab may not be the best purpose of autonomy.
“Building an autonomous robot that moves through the field is interesting,” says Alex Purdy, head of John Deere Labs. “But if it can’t go and execute the jobs like spraying or planting that a grower needs to do in the field it doesn’t achieve what a grower actually needs and is more like a toy.”
And while full autonomy may not exist yet in agriculture, it’s not a new development to the industry — in fact, it’s been around for decades, says Matt Rushing.
“We’ve automated so many things over the years, even though farmers don’t necessarily recognize it as that. Too many people think that autonomy means removing the driver completely,” says the vice president of the Global ATS product line, for AGCO.
For example, going from mechanical to electronically programmable hydraulics, manual steering to auto-steering, and manual headland management to automatic headland management are all examples of autonomy in agriculture.
There’s still a lot of progress to be made before farmers can have their equipment do all the work for them, but companies are making headway with autonomous solutions, which may have profound impacts on both farmers’ and dealers’ operations.
Replacing Tractors, Shrinking Footprints
Kraig Schulz, CEO and president of Autonomous Tractor Corp., says the true definition of autonomy is when a piece of equipment is on its own, driving and functioning completely separately, and all you have to do is observe the results of the work. But that’s not likely to happen anytime soon.
"The world is moving to electric vehicles and I fundmentally believe it's going to have a huge impact on dealers' businesses because the traditional model is slowly going away..."
— Kraig Schulz, Autonomous Tractor Corp.
“Autonomy works really well in repetitive, mundane tasks,” he says. “I don’t think you would get many farmers to say that it’s the same thing every single time they go out; it changes every time. And that’s what makes true autonomy extremely difficult. And not even necessarily the answer.”
Schulz explains that data from Iowa State University Extension shows that for corn following soybeans, labor only makes up 5% of the total cost of production, which means that removing the farmer from the field isn’t going to make a big difference to the profitability of an operation.
Instead, he says the problem with labor in a corn-soybean operation is having the right people on hand at the right time. And that’s where he sees semi-autonomous equipment coming in.
“It doesn’t have to go out while we’re filling up our second cup of coffee and come back. It has to operate independently enough while we’re within a reasonable distance of that vehicle, so that we can take care of all exceptions like we would if we had a driver.
“One might imagine that you wouldn’t actually need as many tractors because you could have self-powered, self-propelled semi-autonomous implements doing the work alongside you in the field during the key seasons where you don’t have enough labor.”
AGCO already has one such solution that is now up to series-production readiness for its Fendt brand, Project Xaver — a swarm of small, autonomous seeding units.
The manufacturer first started the project by thinking about how to remove the driver from the tractor but realized there were still plenty of agronomic and logistic problems to solve, Rushing says, such as soil compaction caused by the tractor weight. It would also be inconvenient to transport.
“We looked at the whole planting process and thought, ‘So what does it take to plant corn?’” Rushing says. “Well, you need a tractor, you need a planter bar, you need the row units that actually hang on the bar and that’s what does the planting. But in reality, what really does the planting? It’s the row unit itself. Everything else is just a necessity to propel the row unit through the field.”
By replacing one 8-row planter and tractor with a dozen or so autonomous row units, Project Xaver not only reduces compaction, but is also designed so that if one row unit goes down, another one can step in to finish the job, reducing downtime. Unveiled at Agritechnica 2017, a swarm of 10 Xaver prototype units can plant about 2.5 acres per hour and are automated to refill and recharge batteries, which takes about a half-hour.
“Years ago we had lots of little machines going through the field and they were all manned,” Rushing says. “Then all of a sudden we went bigger because it was harder to find labor to operate them. But then if we see a significant failure on one big machine, the farmer may have a significant downtime issue.”
Ohio State University ag engineering professor Scott Shearer sees soil compaction being a big driver for smaller autonomous equipment.
“A lot of farmers today don’t realize the compaction penalty as it applies to profitability of their operations,” he explains. “I’m not going to disagree with the fact that farmers buy larger equipment to be timely, and that’s a very important aspect in terms of profitability. But the other side of the coin is, I don’t think they recognize the amount of yield reduction that is occurring because of the larger equipment.”
He points to another company working on this, SwarmFarm Robotics, based in Australia. The company is also moving away from large equipment to swarms of small, autonomous equipment, such as sprayers and mowers.
With these developments, he believes that “supervised autonomy” will be the first adoption of autonomous equipment, where a human is monitoring a fleet of small machines, until artificial intelligence can replace that person.
“We’re going to go from one person sitting on a machine to one person sitting in an office monitoring a machine, to a person in an office maybe monitoring a dozen or two dozen machines,” he says. “And as we learn more about how to give that machine in the field intelligence, we’ll reduce our reliance on the human monitor.”
Automation Before Autonomy
As for the immediate future, Purdy believes more automation has to occur before agriculture starts seeing broad-based autonomy.
“Our group is focusing most of our attention on automating the tasks in the field to give the grower all of the tools to automate and improve the accuracy of important in-cab decisions. Not only will this increase precision and profitability for growers, it will allow for less-skilled labor to be in the cab and be as effective as a skilled laborer,” he says. “We actually think that those automation questions instead of autonomy questions are more difficult to achieve and will be eventually required to get to a full autonomous solution.”
He adds that John Deere Labs, a part of the company’s Intelligent Solutions Group, based in San Franciso, is looking at what they think growers are going to want and need in the next 10-15 years. He believes that automation is the most important and critical piece of the puzzle.
“We’re pretty excited for some of the automotive companies to really drive down some of the costs in full autonomy and drive down some of the challenges we’re going to run into in full autonomy,” he says.
Fuel of the Future
One development the automotive industry is driving with autonomous vehicles that experts believe will make a big impact in agriculture is electric motors.
Schulz, Rushing and Purdy all say that autonomous equipment will likely be electric due to the amount of power the equipment will require, as well as the precision capabilities it will provide.
"If you can manage the liability of slf-driving vehicles in downtown Los Angeles, we can probably learn to manage the liability in the middle of a corn field in Nebraska..."
— Scott Shearer, Ohio State University
For example, Schulz says there are some farmers pulling trains of equipment through their fields and overpowering their tractors. With electricity, those implements could have their own supportive power systems.
“You have greater precision of the implements because you’re directly steering them,” Schulz says. “Think about that from a precision standpoint. No longer are you dragging stuff through the field, you’re carrying it. You can control and power the entire system, creating huge efficiencies on the precision side.”
Rushing adds that in addition to precision control, electric would eliminate a lot of mechanical and service issues as well.
“The need for power is even higher than it used to be,” he says. “So now we will have to either go to bigger alternators, generators and motors and all these other things, or we can focus on making the whole vehicle electric.”
Purdy notes that the artificial intelligence equipment they have from Blue River Technologies, which Deere acquired in September 2017, requires a power source and agrees that electricity and electrification of implements will probably be necessary as we move to smarter equipment.
Moving to electric isn’t without challenges, though, the primary one being the cost of batteries.
“All the systems that are out there for electric cars are based off the premise that you’re going to use a battery,” Schulz says. “They don’t like running off of generators, so we use batteries.”
The problem is the amount of energy a tractor requires is much higher than an electric car. For a 200 horsepower tractor to run for 10 hours would require 1,500 kilowatt hours, Schulz says. That would cost $350,000 in lithium batteries alone and would weigh more than the tractor itself.
Schulz thinks the better option for agriculture is to find a way to couple a generator to the engine and use electric wheel motors without batteries to power the tractor.
But there are agriculture companies introducing battery-powered vehicles. Rushing says AGCO released its E100 tractor that uses a lithium battery, noting that it can run for about 5-6 hours without a recharge and recharges to about 80% within an hour.
“This stuff is coming,” Schulz says. “The world is moving to electric vehicles, and I fundamentally believe it’s going to have a huge impact on dealers’ businesses because the traditional model is slowly going away.”
Changes for Dealers
Part of the reason Schulz believes electric-powered vehicles and equipment will make its way to agriculture — and that it will have a profound impact on dealerships’ business — is due to the simplicity of it.
“I was talking to a friend who bought a Tesla and the dealer was 400 miles away. He said, ‘Well, what am I going back to the dealer for? Why do I need that?’” he says. “That’s a challenge when you start thinking about this. If we start moving toward electric systems and the longevity goes up and simplicity is improved, how does that change your business models? Farmers aren’t coming in for repairs and new parts all the time.”
For example, he explains that a tractor could go from having at least 20 ECUs to less than 5 because it can be monitored and controlled digitally. Several wires could be replaced with one fiber optic cable.
“You’re going to have multiple ones running different places, but you can literally take all the communications you have and boil it down to one line,” Schulz says. “You know how you check that line? You stick a flashlight in one end and if light comes out the other end, it’s good. That’s simplicity.”
Electric motors also last longer. The U.S. Department of Energy’s average expected lifespan for an electric motor greater than 100 horsepower is 29 years, or 200,000 hours of use.
They’re also more efficient. Statistics from the U.S. Energy Information Administration show that going from a gasoline-powered vehicle with a mechanical drivetrain to one that’s gasoline-powered with an electric platform increases fuel efficiency by about 30%, Schulz says.
Because of these advantages, Rushing thinks dealers will start bundling services with the cost of a piece of equipment to make up for the loss of income from parts and repair.
“Most of the revenue generated in a dealership is through their parts and service capabilities,” he says. “So I think you’ll see is new capabilities on a machine, coupled with additional value-added services that you purchase at the point of sale.”
For example, the dealer could offer to include an annual pre-season inspection, load their prescriptions for them or provide additional uptime and technology optimization services.
“Farmers like that,” Rushing says. “It’s like buying an additional insurance policy because now they can depend on their dealer or service provider to ensure the technology’s working, through the service they purchased it and they also got this service at the point of sale.
“There’s going to be more reliance on equipment dealers and technology specialists to make sure the technology is working the way it should, and these services will be offered as part of the purchase.”
Rushing also thinks that as more “swarm” equipment like Project Xaver’s autonomous row units and SwarmFarm Robotics’ autonomous spraying solutions become available, ag service providers may begin offering these units as a service.
While dealers’ parts and repairs business may decrease, Shearer suspects that farmers may trade their equipment out more frequently as new technology continues to render previous technologies obsolete. For example, instead of using equipment for 20,000 hours, a farmer may only use it for 5,000 hours.
“We buy new smartphones because they do more. We don’t necessarily buy new smartphones because the previous one quit working,” he explains. “So I think some alignment of mechanical life and technical obsolescence will be a key consideration when designing.”
Acceptance, Adoption of Autonomous Equipment
A major factor in the advancement of autonomous equipment and other smart technologies is adoption by growers.
Rushing doesn’t think we’ll see every farmer running out to buy something like Project Xaver because it’s a different concept than what they’re used to. But, in time, they will accept it as a tool to solve specific agronomic and logistic problems on the farm.
“You hear farmers say all the time, ‘Hey, I like to drive my machine,’ even though many haven’t been driving for 15 years due to auto-guidance technology,’” he says. “But similarly, farmers in the early 1900s said, ‘How could anything replace my mule?’ But this was then disrupted by early mechanization. Similar adoption reservations and concerns will be seen as we move into the autonomous era.”
Safety regulations will also play a role, but Shearer doesn’t think it will be too big of a challenge.
“There are a lot of companies right now focused on automation in transportation,” he says. “My attitude is if you can manage the liability on public thoroughfares in downtown Los Angeles, we can probably learn to manage the liability in the middle of a corn field in Nebraska.”
While there are companies coming out with their own technology, such as AGCO, or acquiring Silicon Valley-type companies like Deere did with Blue River Technologies, Shearer questions whether some of the immediate autonomous technologies would be applicable to the Midwest.
“I’m not certain that Silicon Valley, if you would, understands Midwestern agriculture,” he says. “That’s not meant to be critical, it’s just a fact of reality. I think they see things at times from a California perspective, and that’s all well and good because there are a lot of high value crops out there.
“But I’m talking about the 200 million acres in the heartland, the corn and soybean type row crops, and that’s going to take a bit of a different mindset. It’s a much tighter margin in terms of profitability, but the acreage is there to support a decent market in terms of machine numbers. So are you going to build a few machines for very specialized markets or are you going to build a lot of machines for the more traditional row crops? It’s going to be interesting to see how some of that unfolds.”