Big Data, Big Yield: How Space is Powering Precision Agriculture
By 2050, you have to produce about 70 percent more food (by cost) to feed the world population — but how do you keep up with that?
That’s what Joshua Miller asked himself when he founded FarmShots. Miller has been thinking about this question for years. “First you have to affect the plant itself,” he explains. The agricultural industry has already been affecting the plant for a long time — developing hearty, pest-resistant breeds, designed to produce high yields under a variety of weather conditions. But there’s a second part of this equation — something other than the plant that has to be optimized. Miller explains: “now we’re moving into an era where we have to monitor and improve the environment around that plant.”
While enrolled at Duke University, Miller (along with a handful of friends) began work on a cutting-edge software product designed to do just that — analyze and optimize the environment around the plant. They called the product “FarmShots” — and now FarmShots has grown into a thriving Precision Agriculture company. Farmers, agronomists, scouts and investors can receive aerial and satellite imagery of their land in FarmShots, and receive a quick and thorough analysis.
Miller drew from his experience as an undergrad intern at a large agribusiness to develop software that delivers fascinating (and actionable) analysis. FarmShots hosts over one million acres of imagery including satellite data from Planet’s satellite constellations.
FarmShots struck an early partnership with Lee Smith of Planet, training FarmShots software with RapidEye satellite imagery, along with other datasets. With recent and archival imagery of fields, both agronomists at a desk and scouts in the field access timely reports, and understand their changing farms right from their computer or phone.
The Data
While the company is only a few years old, it already sports a strong roster of customers.
A large agricultural firm was an enthusiastic first customer. Not long after that, agriculture giant, John Deere, signed on to provide FarmShots as an available add on.
“We’ve helped people determine if farm equipment is functioning properly. If a center-pivot nozzle is clogged, we can use computer vision to detect circles which show that a spout isn’t working.” Miller says.
“We experiment with change detection, or pattern recognition — changes in the image patterns could suggest disease, pests or rot.”
The Future
So how will we create 70% more food in 30 years?
Step one — more data. According to Miller, Planet’s goal of a daily, high-resolution picture of the Earth is a game-changer. “I’m personally very excited by the prospect of daily data. And our customers are even more excited by it. For context, it can take about three to four days for a problem to pop up in a field. If you can monitor your farmland on a daily basis — that’s when farmers will start to get 3-times or 4-times the economic value that FarmShots is already delivering today.”
With a huge, frequently updated satellite imagery catalogue, farmers and agronomists can not only spot anomalies, but rely on historical data analysis to help diagnose them faster.
With the wealth of data collected by Planet and other imagery providers, comes the need for more automated methods of image analysis. According to Miller, a host of data providers, agricultural professionals, and software developers will need to collaborate to create next-gen agricultural analysis methods. “It’s a marriage of agronomic knowledge and computer vision that helps us get closer and closer to automated anomaly detection” he explains.
The Precision Agriculture movement is kicking into high gear. Some of the world’s most creative technologists and knowledgeable agricultural professionals are using the latest technology and working to create healthier farmland, higher yields, and a more sustainable future.
To learn more about the data and software powering smart farms, visit planet.com and farmshots.com.
