SAN FRANCISCO – Capella Space, a startup that builds a constellation of synthetic aperture radar satellites, joins SpaceNet, a non-profit organization focused on geospatial applications for artificial intelligence. SpaceNet was founded in 2016 by CosmiQ Works, the In-Q-Tel organization with a focus on commercial space, and DigitalGlobe, now part of Maxar Technologies.
SpaceNet offers free high-resolution electro-optic satellite imagery, solicits challenges, and awards cash prizes to encourage users to test and optimize image analysis tools.
SpaceNet has helped expand geospatial research and applications for electro-optical satellite imagery, said Ryan Lewis, general manager and senior vice president of In-Q-Tel, the venture capital arm of US intelligence. “Looking to the futurewe see the same not only for images but also for SAR data, “he told SpaceNews.
For Capella, joining SpaceNet is a great way to “bring SAR data into the hands of a broader set of data scientists and software engineers,” said Scott Soenen, vice president of product engineering at Capella, “this is the first step in ourCommitment to the wider community. The publication of data under an open license will encourage the development of non-traditional applications for SAR data. Capella has raised more than $ 50 million for a constellation of 36 SAR satellites that can capture images at a 50-centimeter resolution and retrieve sites every hour.
Capella launched its first satellite in December 2018. The company did not publish data from this spacecraft. Capella plans to launch its first operational spacecraft, Sequoia, in early 2020.
Capella also captures images with his SAR payload in an airplane. We collected aerial data and created it according to our product specifications for our upcoming satellite data, “said Soenen.We will have a high-resolution, hourly data set to contribute to the SpaceNet challenges ahead of Sequoia’s routine operations.
While some people find SAR data to be more difficult to process than electro-optical images, one of the data sets provided is so similar that some of the same computer vision tools are also suitable for electro-optical and SAR data, according to Soenen. SAR satellites also collect images and data through clouds, smoke and at night.
At Capella, we want to overcome these barriers to starting to work with SAR and make sure we offer products that make work easy, “said Soenen.Working with partners like CosmiQ Works and the SpaceNet team, we’ll make sure there’s a good selection of open source tools that can easily connect to our dataset, “said Soenen.
In SpaceNet’s fifth and most recent challenge, competitors automatically identify roads and use travel times to determine optimal routes by applying artificial intelligence to satellite imagery. Since the first challenge in 2016, “we’ve seen the community grow in maturity over time,” said Lewis. “That’s why we think it’s an optimal time to focus on how SAR ina future challenge can be included. ” Prior to Capella, SpaceNet partners were CosmiQ Works, Maxar Technologies, Intel AI, and Amazon Web Services.
The partners have agreed that geospatial mapping problems are rarely solved with a single data type, “said Lewis.While we have seen tremendous improvements in the models we have gained through both our internal and our open sourcing research challenges, we are now ready and the open source communities are ready to take the next step,where two different types of data are used simultaneously to solve a problem. We could not be more excited to explore what it’s like to combine both electro-optic and SAR data to solve a basic mapping problem. According to Lewis, SpaceNet may include SAR in its sixth challenge, which is scheduled to begin in early 2020.
In the past, high-resolution SAR data was largely reserved for government and defense applications, said Capell’s vice president of business development, Andrew Ulmer, in a statement on August 21 announcing the company’s participation in SpaceNetIn order to access this type of data and reduce the barriers to adoption, we want to encourage broad and rapid progress in trade, conservation and well-being in many sectors.
We encourage industry leaders, academics and non-governmental organizations to experiment with the data as it is our collective imagination that releases the greatest value and changes our lives.