The National Reconnaissance Office plans to make awards to radio-frequency sensing satellite providers as soon as next month, as spy agencies move to expand their work with the commercial sector in both space and artificial intelligence capabilities.

Pete Muend, director of the NRO’s Commercial Systems Program Office, says the agency is evaluating proposals requested under the agency’s “Commercial Strategic Enhancement” Broad Agency Announcement.

“We do have proposals, and we’re eagerly looking forward to completing that source selection and awarding multiple awards in all likelihood as soon as next month,” Muend said during an Aug. 25 webinar hosted by the Intelligence and National Security Alliance.

The NRO issued the BAA last fall as a new method to periodically issue rounds of awards for different kinds of commercial space capabilities, giving companies in the commercial sector a chance to begin working with the spy agency under study contracts that could expand to broader partnerships.

The NRO made the first awards under the BAA earlier this year to five commercial satellite radar vendors.

With the radar studies well underway and the radio-frequency sensing awards pending, Muend says the next focus for the BAA will be hyperspectral imagery providers. He said the NRO will issue a call for proposals from hyperspectral companies later this year.

In addition to the BAA awards, the NRO also expanded its primary contract for commercial imagery in May when it awarded three contracts to contracts to Blacksky, Maxar and Planet, respectively. Maxar had previously been the sole provider of such capabilities for nearly a decade. The NRO did not provide exact figures but said the contracts are “valued at billions of dollars over the next decade.”

NRO officials say the agency will “buy what we can, build what we must,” meaning it will still develop and build custom satellites where commercial capabilities can’t fulfill requirements.

“There are still going to be needs that the commercial is not yet able to satisfy or won’t be able to satisfy particular around sensitivity or classification and those sorts of things,” Muend said. “We’ll continue to have to build other national systems to meet those more difficult requirements, more and more challenging requirements, especially as our adversary continues to present more and more difficult challenges for us.”

Geospatial AI, ML programs expanding

With the NRO expanding the sheer amount of imagery in what it refers to as a “hybrid architecture,” however, the agency is also increasingly using artificial intelligence and machine learning to manage and analyze its increasing imagery resources.

“We’re using AI and ML to help drive visualizations and common operating pictures as well, to make sure the warfighters can make the decisions that they need, at the speed they need,” Muend said. “We’re using it to reduce latencies to detect leading targets. Those targets and adversaries are just getting more and more challenging.”

The National Geospatial-Intelligence Agency, meanwhile, is planning to take over management of Project Maven, the Pentagon’s artificial intelligence pathfinder, starting in fiscal 2023.

David Gauthier, NGA’s director of commercial and business operations, says Project Maven’s algorithms will be used against the data the agency ingests, regardless of whether the information comes from classified satellites or commercial sensors.

But he also said NGA’s commercial services program is interested in buying analysis that may use artificial intelligence and machine learning capabilities developed by industry.

The goal is to ultimately bring both the government’s in-house algorithms and commercially developed capabilities together to help with “deluge control,” Gauthier said, referencing the potentially overwhelming amount of geospatial information available to the agency.

“We’re looking for the best-of-breed algorithms to use in both cases, both the ones that are in-house operations, and the ones that are being used in private industry,” Gauthier said. “And what I think is interesting about working with the Maven program going forward is we’re looking at how we also cross the streams. How can we cross-pollinate the best algorithms that may be curated in the government and pass them to industry for getting better solutions to us. And then how do we curate the best algorithms in industry and buy that intellectual property and bring it into the government so that we can use it against our data holdings?”

Copyright © 2024 Federal News Network. All rights reserved. This website is not intended for users located within the European Economic Area.