If you know what to look for, those endless rows of corn that paint the Midwest in summer are full of a lot more than just cattle feed and future Doritos. They’re full of data.
Now, it’s not news that data science can and should be applied to agriculture. The field of precision agriculture has received a lot of attention over the past few years thanks to advances in sensors and computer vision technologies, and Silicon Valley venture capitalists are now lining up to fund startups that can apply the power of predictive models to all that data. Last year, agri-business giant Monsanto bought a San Francisco-based startup called Climate Corporation, which crunched weather, soil and all sorts of other data to power crop-insurance models, for $930 million.
But the story of Mankato, Minn.-based company Farm Intelligence is still interesting. For one, it proves that you don’t need a Silicon Valley connection to make it big in the data business. It also highlights just how much data we’re talking about when it comes to quantifying the farm. Hint: It’s a lot.
The way Farm Intelligence works, according to CTO Steve Kickert, is by working with corn and soybean farmers to help them make better and more timely decisions. It analyzes sensor data, data from other precision agriculture tools, aerial images, government data and weather data to try and figure out what’s going on in the field. It might notice that plants aren’t high enough or some are missing, for example, perhaps suggesting the presence of disease or some environmental stress. It might recognize the telltale signs of aphids.
Like a loudspeaker over the field, Kickert said, “The crop is actually talking to us.” And then Farm Intelligence talks to the farmer via visualizations and alerts telling them what it has found.
If the numbers are any indication, the product, which is delivered as a cloud service and is only a few years old (it’s actually an affiliate of an older company called Superior Edge) seems to work. Farm Intelligence is managing about a million acres of land right now (most of its users have at least 1,000 acres), but Kickert expects that number will be well into the eight-figure range soon enough. And as the technology advances, it’s figuring out ways to capture even more data about each one of those acres.
“We’re getting wider as well as taller, so to speak,” he said.
Already, added Scott Colestock, the company’s director of cloud operations, “we expect to be at petabyte scale at the end of this growing season.”
If the company can expand out of the United States and into, say South America, which produces a large percentage of world’s soybeans, its total acreage and data volume could skyrocket. Kickert stopped short of saying Farm Intelligence will outpace Google should such an expansion happen, but he does think the company could have more data than a lot of other more well-known companies.
That’s why when Kickert joined the company in 2013, one of his first orders of business was to move the company’s infrastructure into the cloud where it could scale at a moment’s notice. Now, all of its computing infrastructure is running on Amazon EC2, but it’s using a provider called Zadara to manage its growing cloud storage infrastructure.
However, while the scale of Farm Intelligence’s operations (and likely the whole field of data-driven agriculture) might be impressive, its underlying mission is the epitome of the knowledge economy. Like everyone from fertility prediction app Ovuline to music data specialist The Echo Nest, it’s taking advantage of easy access to data and cheap (easily outsourced) computing power and storage capacity in order to put information into the hands of people — farmers, application developers, hopeful mothers — who don’t want to bother with any of that.
“Our primary and, frankly, only goal is to help the farmer … increase the yield they’re getting on their crops,” Kickert said. Some of the techniques for doing that have been proven in academia for years, so now it’s just a matter of commercializing them into a product that scale across thousands of individual users.
“We’re not trying to invent the science side of it,” he said, “as much as we are trying to help the farmers access that.”
Feature image courtesy of Shutterstock user rsooll.