Descartes Labs, a software startup using artificial intelligence to process satellite imagery, has raised $5 million in Series A funding in a round led by Cultivian Sandbox, the veteran agtech venture capital fund.
Through a form of artificial intelligence called deep learning, Descartes claims to be able to make agriculture predictions that are more accurate than those made by the US Department of Agriculture.
Computers will be taught to identify specific information and find patterns across vast amounts of image data and pixels to give a better understanding of how crops are growing globally and over time. In a LinkedIn post, Descartes’ CEO Mark Johnson said the company had looked at 2.8 quadrillion pixels of satellite data to predict the corn harvest by effectively measuring the weight of the average kernel of corn to better than 3 milligrams, or 99 percent accuracy.
Crosslink Capital, Data Collective, TenOneTen Ventures, and ValueStream Labs also participated in a round that cements the company’s decision to focus on agriculture.
This is the second round of funding for Descartes Labs within one year; it raised a $3 million seed round in December 2014, which Crosslink Capital led before the company had decided to focus on agriculture. The company soon realized the need for better processing of data for agriculture and the need for a daily understanding of crop production, Johnson told AgFunderNews.
“We didn’t really need the extra funding, but we had made so much progress by the end of the summer that lots of venture funds were getting in touch. We then found a great partner in Cultivian who encouraged us to raise more,” he said. “Cultivian is an incredible ag-focused fund with deep connections in the industry through their LPs and portfolio companies. We’re a bunch of physicists trying to solve some big problems so their experience is really important to us.”
Descartes’s technology was first developed at Los Alamos National Laboratory in New Mexico where the team spent years developing artificial intelligence technology and working on global events like hurricanes and wildfires.
“But we weren’t sure exactly where to start in turning that science into a product or business,” wrote Johnson in a LinkedIn post. “In my career in many science-based startups, one of my observations is that focus separates the good from the great. So, after spending a few months exploring potential projects we decided to focus on data-driven agriculture.”
Andy Ziolkowski, partner at Cultivian, will join the board of Descartes and hopes to find synergies with some of the firm’s existing portfolio companies, which range from farm management data platforms such as Conservis, to soil moisture detection technologies such as AquaSpy.
“We’ve been looking at this whole area for a long time now. We’ve looked at quite a number of satellite, drones, and sensor companies, and we conclude that the key will be the data that they’re generating,” said Andy Ziolkowski, managing director at Cultivian. “The challenge is processing it and we see that as a real bottleneck. What we hope Descartes will do is provide a solution to really take that data and analyze it effectively and quickly.”
Descartes has already entered into a partnership with Data Collective’s portfolio company Planet Labs, which will involve access to the satellite imagery company’s early data, according to Johnson.
“Planet Labs is very complementary to what we’re doing. With their daily pictures of the globe, there’s a just a boatload of data coming down the pipe which will need specialty shops to do that analysis and derive real value from that data. No-one wants pixels; what they want is the data derived from that,” he said.
And what about drones?
“We have a wait and see approach to drones,” said Johnson. “The underlying technology of Descartes Labs has been applied to drone imagery, but we are currently looking at a more global scale. Drones are point solutions for fields; where to put fertilizer down etc. But we do see them as something we will look at some point in the future.”
Descartes’ customers will range hugely from commodities traders and analysts to insurance and reinsurance companies doing risk modelling and loss forecasting, to big agricultural companies and everything from trading to logistics.
“Even governments could be customers,” said Johnson. “One of the astonishing things we’ve learned is there are a few countries which really understand how much food they are growing, and really the US is by far the best. Everywhere else the numbers are not very good. I think there’s a real opportunity to give better visibility to countries about what is actually growing there.”
Have news or tips? Email [email protected]
Image credit: Chrishonduras
— Check AgFunder for agriculture investment opportunities —
