The biggest misconception that investors have about the roboticization of agriculture is that farmers are unusually cautious about new technology. It’s not true, and that myth is leading too many in the robotics industry to miss the obvious.
Crop farmers, like all successful business owners, need to be convinced of their return on investment.
“They’re doing cost-benefit analysis,” said Dale Moore, executive director for public policy at the American Farm Bureau Federation, the same way they do for any other investment. “I don’t think they’re necessarily more cautious or superstitious about robotics.”
In fact, precision guidance and autonomous milking, both robotics-related technologies that already enjoy growing customer bases, have caught on because they’ve proven they can provide an ROI, Moore said.
Food producers have adopted technology at roughly the same pace that white-collar businesses switched to desktop computers, for example.
Like any successful farmer, Dan Phelan, an established crop and livestock producer who manages 300 acres outside Wilmington, Ill., closely follows trends in farm technology and vehicles, and is undaunted by the idea of autonomous tractors.
“But, my God, how many acres would you have to manage to cover the outlay?” he asked rhetorically.
Multiple U.S. manufacturers have trials, experiments, prototypes and even a few signed leases, but no one product, no matter how slick, is going to entice enough farmers out of the stands and onto the field.
Investors should know
Assuming that robotics will pull more profit out of agriculture (and it will), here are what farmers will want manufacturers to demonstrate:
- Turnkey products
- Local, competent robotics service
- Bulletproof 24-hour support (Tractor dealers will want this too)
- Systems-management software that’s dead simple
- Robust tractor-level firmware
- Actual autonomy — not autonomy until waves of amber grain freak a system out
- A superior degree of indigenous coding savvy or ties to a solid software partner
- A non-proprietary operating system
- Updateable — often customized — application software
That’s a top-down redesign of the tractor-making business, even if you only consider products that retrofit conventional tractors.
Any company talking about an exciting little innovation unit that it has working on something cool should fall from an investor’s radar like a meteor.
Nor should investment dollars necessarily be wooed by robotics for immense machinery. Tractors have gotten massive to increase productivity, but robotics at any scale is a productivity-enhancer. Each system requires the attention of something much less than a full-time-equivalent employee
At the same time, creating huge robots will limit systems to Big Agriculture when most farming in the United States is still a family-run operation. It also will miss the flexibility and efficiency that swarms of robotics can offer.
Two to watch
At this time, no domestic maker of farm implements has all the pieces to be a serious contender. There are two players to watch, however, and each sits on opposite ends of the tractor industry.
The first is Deere & Co., whose founder, Illinois blacksmith John Deere, made his first farm implement in 1837. Deere reported earnings of $3.065 billion in fiscal 2012.
The second firm is Kinze Manufacturing Inc., founded as a welding shop in 1965 by 21-year old Jon Kinzenbaw in Iowa. Now run by his daughter, Susanne Veatch, Kinze is a comparatively small (850 employees) and privately owned farm-implement maker.
Kinze: nimble and noteworthy
Of the two, Kinze is more interesting if only because it’s far more open about its robotics development than goliath Deere. Kinze has demonstrated both a robotic planter and a tractor/grain cart combination, which is used to take grain from a harvester in the field and deliver it to a waiting truck.
There are conflicting reports about possible leases signed in 2012, a year after the platform was first announced. The grain cart, which Kinze is promoting fairly heavily, consists of:
- Wi-fi/cell communications between a stock manned harvester and stock tractor/grain cart
- A four-button control panel and GPS for the harvester’s driver
- Firmware on a tractor and on the Kinze grain cart that it pulls
- A stock, adaptive cruise-control radar
- A camera
Drivers can command the autonomous grain cart to approach and pace the harvester (so it can accept grain as the harvester reaps), idle after getting grain and return to its unloading post. Here’s a video of the process.
Rhett Schildroth, Kinze’s product manager, said, “We designed the system to plug into any brand of tractor.” Online videos on Kinze’s site show a John Deere tractor pulling a Kinze grain cart.
Kinze’s strengths are many. As a medium-sized business, it’s more nimble, and it has fewer product lines to integrate with autonomous systems. Also, margins on robotics systems likely will be higher than on “big iron,” and, again because of Kinze’s size, harder to ignore than would be the case in a huge multinational.
At this point, the company doesn’t have to sell an overarching computing and communications infrastructure, because its systems are controlled in the field. It can afford to wait for industry standards to be created and plug into a future enterprise-resource-planning package for automated farming.
Less clear is how Kinze’s marketing strategy will play.
“We’re focused on selling farmers greater efficiency,” said Schildroth. “We don’t feel it needs to be bundled” with any other products. There is some sentiment in the market that manufacturers will get more buy-in by blending the cost of new systems with the price of new, greener and pricier engines, for example.
Kinze’s chief problem is one of its assets: its size. Schildroth said, “The market is demanding more than we can serve.” If that’s true, the multinationals easily could step in before Kinze can gain real traction.
John Deere: concerns over systems support
Scale also is a challenge and opportunity for Deere. It can’t turn on a dime, but every degree it does turn could swamp smaller players.
Deere is circumspect about its plans for autonomous systems. Asked for even a roadmap for the company’s robotics efforts, a spokesman pointed to Deere.com. Told there was little information there, he chuckled and said, “Well, that’s what we have to say.”
A video of Stewart Moorehead, manager of Deere’s robotic systems, and posted in March on agprofessional.com, acknowledges a research program on autonomous orchard work.
On video, Moorehead says most of the technologies being used in that program “are quite highly transferrable” to “broad-acre crops” like grains. Deere’s early robotics, he says, are integrated with existing products.
Deere’s challenges include technical issues, cost concerns and the whole matter of service and support, says Moorehead. It’s that last bit that particularly deserves an investor’s attention.
Deere & Co. is a global manufacturer and one supposes that corporate executives could dictate training for and adoption of robotics product lines. If a company with this much electronic and physical infrastructure worries about service and support, how close is autonomous farming?
John Reid, director of enterprise innovation and technology for Deere, said the farm-tractor industry is still getting its head around autonomous vehicles.
“I used to think we [tractor makers] would lead automotive into autonomous vehicles,” said Reid. Farming is dangerous, but not as perilous as an urban freeway.
He remains optimistic, though. Acceptance on the roads will open minds in the field.
In the meantime, investors need to watch for an elusive combination of infrastructure, scale and sharply focused products.