Wade Barnes says he knows what the agronomy of tomorrow looks like. It’s a proactive system that uses data to model crop development, helping farmers make decisions every step of the way.
The power of data analytics will fuel every step, from what variety to plant based on soil moisture, disease and pest pressure and long-term forecasts, through to how to manage the harvest.
Paraphrasing hockey great Wayne Gretzky, he puts it in simple terms: “We’ll go where the puck is going, not where it is just now.”
Barnes’ agriculture journey began in Pilot Mound, where he founded Farmers Edge along with fellow agronomist Curtis MacKinnon in 2005. At the time, the company was looking to jump on to exciting new precision agriculture technology such as variable-rate application based on prescription maps. While Farmers Edge might have been one of the first companies doing this, it certainly wasn’t the only one. But it was in these early days that Barnes says he got his first tantalizing glimpse of the power and potential of information in agriculture.
It happened in Russia, which was rapidly modernizing from the ossified era of the Soviet collective farms, where Farmers Edge wound up providing farms with agronomic advice.
“One of the things that really kind of enlightened me was the size and scope of the farms, and how these organizations made decisions,” Barnes said.
The work in Russia opened another door when he was approached by one of the major life-science companies about helping it source canola and sunflowers for its line of specialty oils.
“They reached out to me as the token Canadian working over there and said, ‘can you help us with this?’” Barnes said.
Suddenly he found himself in meetings between the company and various food companies it was working with. Unfamiliar new terms like ‘traceability’ and ‘carbon footprint’ — common today but rare back then — were suddenly being tossed around.
In the middle of one of those meetings, a Canadian executive of the global life-science company turned to him and said, “Wade, this is where the world is going. It’s not in North America yet, and won’t be for a while, but it’s where it’s going.”
A few years later, another turning point came when the company began flirting with Silicon Valley investors steeped in the world of data, analytics and algorithms.
Barnes wasn’t sold on the idea, at least not initially.
“I actually remember kind of arguing with them that this wouldn’t work in agriculture,” Barnes said.
But he says that on the flight home, he realized they were right, and that meant the whole company needed to change its focus. Farmers Edge would no longer be a precision agriculture company. It would be a digital agriculture company.
Understanding the concepts behind digital agriculture can be challenging in a sector like agriculture that deals in boots on the ground and seed in the soil.
Mike Duncan, an industrial research chair at Ontario’s Niagara College, breaks down just how it works.
It all starts with data — lots and lots of data. Then it moves to analyzing that data and learning from it. The true learning from this is increasingly done by machines — not nuts-and-bolts machines, but software-based artificial-intelligence (AI) technology such as Neural-Turing Machines, named after early computer pioneer Alan Turing. Probably its most well-known use has been in facial-recognition software.
“You take a dataset, train it so it recognizes faces,” Duncan said during a recent conversation. “These are ultra-complex machines, completely hidden away from anyone.”
He said these digital machines can take a vast swath of data and begin to make sense of it. Take the typical corn crop where he lives in southern Ontario. From the data, the machine learns about the plants, their variability and the variation in their entire environment. “It learns where corn grows. It learns that it grows in these areas with this weather.”
Then it starts to pick apart that data and look for patterns. If it sounds complex, that’s because it is, and it relies on a new breed of data scientist.
But these tools are necessary for the massive amount of data being accrued, more than a person could possibly go through and make sense of.
Take that corn crop for example.
When Duncan started working on one research project, he accessed weather data for the region.
“It was weather data at one-hour resolution over 18 years, or 160,000 samples per grid point — and there were millions of grid points. When we got our first block of data it was so big the tools didn’t exist to analyze it.”
In the field
While an academic like Duncan might be excited at the prospect of a pile of data that size, for the farmer in the field, it isn’t even of academic interest.
What they want is better, tightly focused information to make their many decisions just a bit more accurately.
That’s where Curtis MacRae, who farms near St. Andrews, sees the value in working with Farmers Edge as he takes his farm digital.
“The hardest thing about farming is making decisions, but at the end of the day you have to make them,” MacRae said. “You can continue to guess at them, or you can start making more educated guesses. Limiting the amount of guessing lets me sleep better at night.”
MacRae describes it as an ongoing battle where he works to keep incrementally making better decisions, and says those better decisions over time can translate into winning at the game of farming.
“Ag is a long game. Nobody can do it in two years.”
Sometimes the results are counterintuitive. Precision agriculture is often about maximizing production, but digital agriculture is about seeing the bigger picture. For example, based on analysis of several past seasons, MacRae has taken 100 acres of his farm out of production, at least temporarily.
“These are acres I have been continuously losing money on,” MacRae said. “I’ve now seen the data, year after year. Would I see that if I wasn’t using digital? Maybe. But it becomes really obvious and glaring looking at the data.”
In a business of razor-thin margins, every acre has to pay, and that’s a trend that’s not going away, MacRae said.
But that land might not be out of production forever. Up next is a plan to put tile drainage on a portion to see if that nudges it back into the profitable column.
It’s this type of practical application where MacRae sees the value in Farmers Edge and its digital approach. Having a company that understands agriculture on the other end of the equation is important.
He puts it this way: Silicon Valley might understand data science, but it’s used to a stream of 1s and 0s, binary choices and something is either working or not working. But agriculture trades in the grey areas of informed guesses, uncontrollable variables and risk mitigation.
“There’s not many industries that have as much guesswork as agriculture does,” MacRae said. “Everyone outside of farming doesn’t understand the risk until they try and figure out what we do for a living.”
It’s that understanding of the sector that Barnes says sets Farmers Edge apart from the pack.
An early-venture capital investor came to him with an assessment of one of its agriculture advisers: Farmers Edge was a nice little precision agriculture company that would never be able to play in the tech space because “it isn’t who they are.”
Barnes’ response was swift, and maybe a bit cocky.
“Yeah, they have 300 data scientists,” he said. “But I’m going to have better data. And better data will beat data scientists.”
He says that’s a bit of “farm boy logic” rather than “PhD thinking.”
But even at that, he says it’s been a long road to convince people steeped in the technology sector that a farm boy can bring anything to the table.
“I would say most were looking at us and thinking ‘we’ll invest in this company and just get rid of this farm kid and put a Silicon Valley guy in and they’ll be able to move this faster,’” he said.
But once the funding to take the digital agriculture plunge was secured, the company took a path that the tech investors might not have understood.
Most of the big tech success stories are using something that’s known as a SaaS model — software as a service. Probably the best known is Facebook. It’s a digital platform, but relies on every user having their own digital device and internet connection that they pay for.
The other major model is IoT — or internet of things — which requires a lot more investment in digital infrastructure to gather that data. But when it comes to digital agriculture, Barnes is convinced it’s the winning model.
“Silicon Valley, they thought farmers are dinosaurs, and they could make this huge disruption in agriculture,” Barnes said. “But they didn’t really appreciate the lack of infrastructure. Not the lack of intelligence or the lack of technology, but the lack of infrastructure to go out and make an impact in agriculture.
“We said if we’re going to go out to a farm, get that data and analyze that data and help a farmer make a decision with it, we have to get the very best data. That meant going out and connecting all those machines.”
It was expensive, and it took time. But now Barnes says the true potential is starting to show as more acres and seasons of data have accumulated.
“You need to get all this data, get it passively into the cloud, process it, and serve it back to the farmer in a way that makes it easy for him. If you can’t do that, it doesn’t work. But if you can — man, it becomes powerful. I think we’re just kind of getting there now.”
While a digitized grower base of 20 million acres worldwide might seem huge, Farmers Edge and companies like it are still small potatoes compared to the tech titans they’re going head to head with, Duncan says.
Duncan says he had the chance to talk to a senior executive with X — formerly Google X — the ‘moonshot’ research group tasked with finding the next big thing. He asked what their annual budget was, and was told it was unlimited. When he pressed the executive a bit further he was told they’d spent US$22 billion on various projects the previous year.
With that kind of money on the table, smaller companies are at risk of being gobbled up, or simply sidelined. Give a good AI access to enough data and Duncan says the next layer of Google Maps could be Google Farms, generating detailed prescription maps based on data every spring.
“Wade is not a small player in this field, but compared to Google, he’s nothing,” Duncan said.
It’s not the first time Barnes has faced skepticism and it won’t be the last.
“At times it was kind of lonely, going in the direction of this vision, while everyone said ‘you’re nuts,’” he said with a laugh. But, “it’s already working, it’s already happening.”
As for whether a farm kid from Manitoba can take on the tech giants, stay tuned.
This article first appeared in GrowPro, Glacier FarmMedia’s magazine for agronomists and agrologists.