How One Company Aims to Model the World—and All Its Activity—in the Cloud


Descartes Labs is building a platform to let the Pentagon analyze all the geospatial intelligence, all the time.

The amount of geospatial intelligence available today is hitting at an all-time high, but aggregating and analyzing all that information can be a difficult, costly endeavor.

“The job of going to multiple places, contracting and buying and pulling and preparing all this data has really inhibited the deployment of what are really pretty powerful [analytics tools],” said Adam Smith, co-founder of Descartes Labs, in a conversation with Nextgov.

Today’s geospatial intelligence ecosystem can’t keep up with growing demands from the government and private sector, Smith said. With funding from the Pentagon, his company is building a platform that realigns the process for modern-day needs.

Descartes Labs on Wednesday announced it won a $7.2 million contract from the Defense Advanced Research Projects Agency to participate in Geospatial Cloud Analytics program. Through the effort, DARPA plans to aggregate troves of open-source and commercial satellite data on a single platform, as well as create tools to build and test predictive models on that platform.

Under the first phase of the contract, Descartes Labs will work to unify some 75 datasets of real-time satellite imagery, weather patterns, green energy operations, shipping and other areas into a single user-friendly system, Smith said. In the second phase, selected teams will test the platform's worth in three categories: famine, spotting illegal fishing activities and analyzing global fracking operations.

At the end of the 18-month program, various agencies across government will have the opportunity to examine and assess the value of different platforms and tools, according to Smith.

Descartes Labs last month launched a platform that allows users to analyze multiple layers of geospatial data, but Smith said the system it’s developing for DARPA will be “a major step up.” The company sometimes crunches up to 100 terabytes per day to keep its current system up to date, he said, and that’s only a fraction of what it hopes to include in the future platform.

DARPA’s program only focuses on a handful of areas, but the company’s platform will include information “that would enable all of the most potent use cases for the intelligence community,” Smith said.

Geospatial intelligence has played a key role in giving the U.S. a leg up on its foreign adversaries in the past, Smith said, but in an era when any group can access that data, the country will lose its competitive edge if it doesn’t innovate its toolkit.

He sees rapid, multi-layer analytics as that extra advantage for not only government but industry as well.

“Supply-chain [companies] that operate out there in the real world—whether it’s in agriculture, energy, logistics, metals, mining, chemicals, you name it—are going to find their competitive edge eroding if they don’t have access to this data,” he said. “Those companies are either going to have to get onboard with that or they’re going to lose.”