USGIF GEOINT Community Forum Aims to Deliver Information “At the Speed of Need”

On Monday November 16th, 2020 the opening Keynote Address of the USGIF GEOINT Community Forum was presented by Stacey Dixon, Ph. D., Deputy Director, National Geospatial-Intelligence Agency (NGA). The theme of the conference “The Convergence of Commercial Content with AI/ML to Provide Clarity” shaped the narrative for the week ahead. Dixon said that Covid-19  provided an unexpected opportunity to reimagine their mission very quickly.

Stacey Dixon, Ph. D., Deputy Director, National Geospatial-Intelligence Agency (NGA)

GEOINT is a discipline that will ensure geospatial superiority for generations, Dixon said.

“I will set the stage from a strategic perspective,” said Dixon. “As we look at national intelligence strategy, we look at national security strategy, national defense strategy and then look at national intelligence strategy. This summer we wrote the NGA Strategy 2025 as well as our strategy for a national system for geospatial intelligence strategy for 2021-2025. What ties this all together is we developed according to our director’s intent. We find ourselves in an historic inflection point, the nation, our agency and GEOINT community. We are in a time of great change and challenge as well as great hope and opportunity. There is great power and competition testing our prosperity and security. At the same time new technologies and commercial capabilities are redefining GEOINT. Our great power competitors are trying to press their advantage in every domain, not just with sophisticated weapons, but they are pursuing technology to make themselves better than us at GEOINT.”

Dixon noted that at the same the Covid-19 virus has accelerated the rate of change in the strategic environment, and presents new challenges and opportunities. “We will take every opportunity to sustain our GEOINT advantage so we can hold at risk the strategic forces our adversaries use to project power and threaten the U.S. and our allies. In order to sustain GEOINT, we need to do three things: maintain primacy of our core mission, be the world’s premier GEOINT force, and relentlessly pursue a whole enterprise approach which is where USGIF comes in. Our strategy remains focused on people, partnerships and our mission. We continue to emphasize people first mission always. We invest in maintaining and recruiting a world class workforce, and that advances our tradecraft and finds innovative solutions to meet our mission needs. Our goal is facilitate the transformation of our analytical workforce into a data literate technology adjacent workforce of the future that partners every step of the way with our data savvy and technologically adept workforce. So the analysts themselves don’t have to understand and be practitioners of all the data science, but they need to be able to work in close proximity with those who are.”

Dixon said that the NGA invests smartly and pursues collaboration working in close proximity with international partners in order to fulfill their common vision as a resilient GEOINT community.

Vice Admiral Robert Sharp, seventh Director of the NGA, especially likes to use the Moonshot strategy of holding a goal much like President Kennedy did in 1972, holding ourselves accountable to be the best, said Dixon. This type of collaboration and sharing of ideas that allows us to imagine or reimagine our way ahead. “NGA is reimagining the entire GEOINT operation to deliver information faster at the speed of need. It’s an exciting time of discovery and curiosity, creating a fountain of new and game changing ideas. The Moonshot is intended to take all the capabilities that we know we need to develop to maintain GEOINT supremacy.”

“As part of our response to the pandemic, our agencies have embraced commercial services in some unprecedented ways,” said Dixon. “The pandemic really channeled our analysts’ drive and innovation into finding new ways of supporting our mission and customers from home. Covid-19 accelerated the commercial use and exploration of services in a telework environment. The ability to continue to conduct impactful and insightful analysis in a telework environment may be the biggest cultural function our agency has experienced in more than a decade. It gave us the opportunity to partner with industry in new exciting ways. Leveraging commercial data creates superb opportunities for us to further ensure how we partner with industry and how industry partners with us.”

Furthermore, the new opportunities let NGA explore new missions that are critical to NGA. It has energized beyond looking at more imagery partners and seeking more data and analytical service partners. “We have been able to do priority analytic work on unclassified networks,” said Dixon. “It was a great opportunity for workforces to reimagine their mission areas because using commercial data the way we did, on a variety of missions, wasn’t even thought to be possible before.”

There are 90 commercial earth observation constellation satellites covering various imaging modalities like multispectral, hyperspectral, SAR, all around the world. “We use this to develop analysis fully with unclassified data,” said Dixon. “In other cases we start with unclassified information to get these projects going and add on classified data later on and finish that analytic work. For the recent wildfires on the west coast, NGA supported frontline firefighters by providing key geospatial data and collaborating with organizations like FEMA, US Geological Survey and the Coast Guard. We helped in detecting and monitoring new and emerging wildfires in California, Colorado, Oregon and Washington.”

In one specific example, NGA scientists used commercial data to detect and map burn scars, providing a comprehensive view of select burn areas east of Salem, Oregon. They could then quickly send this information to FEMA in under 24 hours. “This is an example of how commercial data helps NGA support our customers faster and easily through commercial channels.”

Another NGA way of putting commercial data to use is through their data labeling project. A group of 600 volunteers from across the agency and enterprise were able to label 1 million objects from commercial imagery chips in just 107 days. These objects are fuel for training algorithms to recognize objects the NGA deems important, like planes, trains and automobiles. Most AI machine learning algorithms need a minimum of 10,000 label objects in diversity sets to operate effectively.

“This was done completely while teleworking,” said Dixon. “With this accomplishment we’re getting closer to NGA’s Moonshot by delivering speed and accuracy to NGA GEOINT and warfighters.”

Commercial data has created innovation in NGA’s work processes. NGA is at a unique intersection between the IC of the intelligence community and the Department of Defense.  Part of their mission is to support humanitarian disaster, provide humanitarian assistance and support disaster relief and also provide safety of navigation both for aeronautical and maritime domains.

“Our mission requires that we succeed at all security levels, and in a public way that many other organizations can’t or don’t,” Dixon said. “Innovation always been part of our culture, regardless of Covid-19, the commercial data, with more satellites in the sky, it has helped shift us even more from imagery and intelligence to activity based intelligence (ABI).

Collection from satellites can focus more on when events occur, instead of mainly collecting against fixed locations which is an incredible game changer.

Dixon cited a recent example of combining commercial data and innovation in the response to the tragic explosion at the Port of Beirut in August:

“We had a team of NASA Fire Information for Resource Management System (FIRMS). FIRMS detects and disseminates information on fire events such as explosions within 3 hours of detection. The team applied commercial modeling to the FIRMs data to detect an alert to the team of when a large explosion occurred. It also centralized automated intelligence collection that helps collection managers and analysts gain improved situational awareness and get the information they need faster, which improves the timeliness of our response.

The team then went on to overlay information on buildings in the area from NSG Open Mapping or NOMEs and OpenStreetMaps over Google, to better understand the features of the explosion site. This made it possible to carry out more of our mission goals of verifying and understanding more about world events. It was an unclassified solution for an unclassified problem.”

This capability became even more important as the NGA had to keep their workforce safe and out of their facilities during the early days of the pandemic. This approach is always vital to being able to support the Humanitarian Disaster Relief Mission.

In another example of an industry partnership using commercial data, NGA collaborated with Amazon Web Services to create Amazon Textract, a machine learning model that automatically extracts tables from aeronautical information publications. NGA employees are responsible for manually identifying changes in these publications collected from over 200 foreign nations and aviation organizations. They average over 250,000 pages of updates every month.

This is for the safety of aeronautical navigation missions and is disseminated every 28 days. Using a trained test to evaluate a repeat method, the model now automates over half the manual review of one country’s publications. After only four months of training this was accomplished during the Covid-19 crisis.

New and old issues can be addressed with the newly developed technology. Great power competitions have re-emerged as a central challenge to national security.

“You’ve all seen the strategic shift in the national intelligence strategy and national defense strategy that places less emphasis on counter terrorism and counter insurgency operations and more emphasis on great power competition,” said Dixon. “Our quote from the National Defense Strategy: ‘Inter-state strategic competition, not terrorism, is now the primary concern in U.S. national security. it is increasingly clear that China and Russia want to shape a world consistent with their authoritarian model. With openly and publicly available data like commercial data, more eyes are on the world and more nations can hold countries accountable for their actions.’”

Back in August, a satellite image from Planet made the news, that was first posted on the social media counsel for Radio Free Asia. It showed what appeared to be a nuclear power attack submarine entering its home, in an underground berth of a UN naval base on Hainan Island off the coast of China. “You may remember, that’s where a U.S. Navy reconnaissance plane needed to make an emergency landing back in 2001 after a Chinese fighter plane collided with it,” Dixon pointed out.

In August, another image popped up on the American Cable News that showed what appeared to be a vessel conducting a live fire of its weapons system over the East China Sea. With all hands on deck, the companies producing those images, plus those observing them and the people who are bringing attention to what they see, national security was greatly improved with this application of commercial data.

The NGA’s technology strategy is published on their website and is publicly available, so that academia and industry can both help them meet their challenges.

Dixon said the NGA are using basic predictive models, and also need to develop mission focused models for analysts to test and create analytical hypotheses and conduct complex participatory analyses. “That way we can continue to be proactive, and not reactive. There is a deluge of data from an ever growing number of sources to inform our analysis. We’re entering an era where we’re no longer thinking we’re lucky to get the image or shot we needed,” Dixon explained. “Now it will be instead finding out who got the image or shot that we needed when we needed it. We will need an effective way to store, organize and manage all that information or it just isn’t as useful. It will be overwhelming if we can’t manage it. With data management tools that operate on a global scale and use data synthesis and visualization, we can actually derive intelligence from all the information we receive. AI is a great tool to apply to large amounts of data.”

Analysts in the future will be able to spend less time sifting through data and more time investigating pertinent data. From industry and academia, the NGA would like to see more companies offering tested and validated market ready analytics that fuse multiple sources of commercially derived and commercially available data to meet NGA’s needs.

What are the NGA’s needs? Target monitoring and activity learning to help maintain their advantage.

“We also need software engineering that’s in a robust development program, a platform that uses continuous international, continuous delivery processes and shared development security and secured operations or dev ops tools,” Dixon said. “Finally, we’re looking for capabilities that make collaboration easier, between people and machines, capabilities that increase the efficiency and abilities of NGA’s internal elements to work together across classifications. And capabilities that will improve and diversify the intelligence products we can provide to our customers and partners.”

Dr. Fei-Fei Li, co-director of Stanford Institute for Human-Centered Artificial Intelligence, said that when it comes to AI, she is an optimist and Dixon shares that view. “I often tell my students not to be misled by the name Artificial Intelligence,” said Dixon. “There is nothing artificial about it. AI is made by humans, and intended to behave by humans, and ultimately to impact human lives and human society.

I’m personally excited about where we’re going with AI machine learning and the possibility of what commercial data can help us do as we incorporate it into the datasets we’re using. We’re on the cusp of great things.”

Tags: cloud, data, forestry, geospatial, GIS, GPS, imagery, indoor mapping, Infrastructure, intelligence, LiDAR, location, mapping, maps, mobile, mobile mapping, NASA, National Geospatial Intelligence Agency, remote sensing, satellite imagery, social media, underground mapping

Categories: 3D Cities, aircraft tracking, analytics, asset management, Big Data, cloud, cloud network analytics, data, developers, disaster relief, drones, emergency response, field GIS, geocoding, GEOINT, geospatial, geotechnical, GIS, government, GPS, handhelds, hardware, lidar, location based services, location intelligence, mapping, mobile, NASA, National Geospatial Intelligence Agency, Open Source, remote sensing, satellite based tracking, satellite imagery, space-based flight tracking, spatial data, SPOT 7 satellite, storm surge, subsurface utilities, survey, telecommunications, transportation, USGS, utilities, wildfire risk, wireless networks

This entry was posted on Friday, November 20th, 2020 at 5:59 pm. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.