AI and Open Source Data Management Disrupt the Farm Industry

With artificial intelligence and open source data management, farmers can better learn what to plant and when, how to maximize soil nutrients and track carbon content, and more.

By Anne Miller, Contributor

To map every gene in the human body, scientists around the world collaborated for more than a decade, from 1990 to 2003. Thanks to their work, entire vistas of medicine have opened up, from new diagnoses to drug regimens tailored to an individual’s genetic makeup. What if, posits Dorn Cox, a produce farmer in New Hampshire, the same could be done for the world’s soil?


With detailed knowledge of the nutrients in their soil, farmers could better tend their dirt and significantly reduce negative environmental impacts. For example, they could better learn what to plant and when, or how to maximize soil nutrients and track carbon content (more carbon in the soil means less carbon in the atmosphere). A new platform, Open Technology Ecosystem for Agricultural Management (OpenTeam), strives to do just that.

More than a dozen major agricultural players, including household names like Stonyfield Farms and General Mills, are investing in OpenTeam, along with the U.S. Department of Agriculture (USDA). Most of the OpenTeam components exist already, such as data platforms for tracking individual farm trends, carbon testing tools for the soil, and predictive analysis.

But before OpenTeam, those tools couldn’t work together nor get as granular, and researchers could only access limited pools of data. OpenTeam’s goal is to bring all that data together in one platform with a user-friendly interface, allowing farmers to make better decisions, and companies and researchers to better track trends.

Cows May Lead the Way

Stonyfield Farms, the national organic dairy company, hesitates to ask their 250 dairy suppliers in the Northeast to make major operational changes without proof those changes will pay off. But if they can make suggestions to their farmers based on detailed ground data from each farm, that could be an environmental game-changer, explains Britt Lundgren, Stonyfield’s director of organic and sustainable agriculture, who helped spearhead the OpenTeam project.

“Every time you’re asking a farm to make changes, you’re asking them to spend money…We want to be really certain we’re asking them to do the right thing.”

—Britt Lundgren, director of organic and sustainable agriculture, Stonyfield Farm

“We’ve looked at different models; we’ve looked at different feeding strategies and different manure management,” she says. “The thing we kept running up against with those efforts is that it was very labor-intensive to collect the data we needed from those farms and hard to assess what it meant for the individual farms.”

“Every time you’re asking a farm to make changes, you’re asking them to spend money,” she continues. “We’ve been afraid to ask farmers to do something and then learn down the road it wasn’t going to work for some reason or another. We want to be really certain we’re asking them to do the right thing.”

OpenTeam allows farmers to easily test their soil via inexpensive, portable tools (which extension programs or the USDA staff may also have), upload data from those tools into the platform via user-friendly interfaces, and have an artificial intelligence (AI) system make suggestions on best practices tailored to a specific plot of land.

For example, many dairy farms use a single type of grass seed and rotate cattle through different areas of the farm to graze depending on grass growth patterns. OpenTeam could suggest additional types of grass that would increase the amount of carbon dioxide and nutrients in the earth; that, in turn, encourages a growth pattern that better supports grazing rotations with less effort from the farmer, healthier cows thanks to more nutrients, and increased milk production.

That data also offers researchers a better bird’s eye view of overall soil and farm health. “Previously, all of this data had to be collected in interviews with farmers that would take several hours each and require multiple rounds of follow-up, even though much of this data had already been tracked by these farmers in other ways,” says Lundegren. “With OpenTeam, a farmer only has to track a piece of data once and it becomes available and usable across the platform, and it can be supplemented by remote sensing data, as well.” For example, satellite weather technology feeds into the system along with those on-the-ground tools, for better predictions about what may occur on the individual farm level.

The $10 million project, funded mostly through grants and corporate partnerships, started development in late 2016, with farms testing and adopting the system over the following three years. OpenTeam publicly launched in the fall of 2019 at Wolfe’s Neck in Maine.

Making the Invisible Visible

Cox also runs a 300-acre, grass-based dairy farm in Freeport, Maine, a supplier for Stonyfield. He also manages a 600-acre farming science lab and created an open source data management system for farmers called FarmOS, a part of OpenTeam.

“Technology is a chance to democratize access to markets. OpenTeam is making all the [invisible] parts of agriculture that are valuable to society … visible.”

—Dorn Cox, farmer and FarmOS creator

Cox’s FarmOS tracks data about a farm, including a mapping component and data from equipment, like a spectrum analysis tool that fits into a pocket and measures soil’s carbon content.

“Technology is a chance to democratize access to markets,” he says. “OpenTeam is making all the [invisible] parts of agriculture that are valuable to society … visible.”

For instance, one of OpenTeam’s technologies is Cool Farm Tools, which allows farmers to input information such as crop types and fertilizer usage, and obtain data about greenhouse gases on their farm, as well as its biodiversity. The tool can track farm changes and award “points” to farmers who make changes to increase biodiversity (the birds and other fauna on their property) or sequester more carbon (removing it from the atmosphere).

These are moves that aren’t typically part of the public awareness when it comes to farming, Cox says, but numbers on paper can demonstrate that farmers are making valuable contributions to the environment.

“At the farm level, we have access to tools to measure things in the field and how we can improve soil health,” he says. “That same data, when combined with satellite imagery, can relate what can happen at one farm to its neighbors, to its watershed, to the state, and the national and global effect.”

“It really transforms how we can talk about these systems and the role of agriculture in influencing some of those environmental systems.”

Lundgren agrees. Stonyfield has a goal to reduce emissions 30 percent by 2030, but she says the brand has already seen disruption due to climate change hit its supply chain. She sees OpenTeam’s potential for change at the individual farm level.

“The hope is that [farmers] are really driving making changes at the farm level,” she says. “We really see this platform as being the underpinning for any ecosystem management,” such as tracking usage to offer carbon or water credits in the future.

Of course, there are hurdles, among them adoption from farmers. Lundgren says Stonyfield would consider an incentive program for their suppliers to engage in OpenTeam, to help farmers, the environment, and their brand.

As Cox notes, data analytics has disrupted industries around the world—so why not farming? It may arguably be the world’s oldest industry, but that doesn’t mean new technology can’t help innovate.