Automation and machine learning become more sophisticated on a daily basis. Pervasive sensors, collaborative mapping and crowdsourcing are creating new global datasets. How can we leverage all this data to anticipate where and when events will happen to save lives, resources and time? This was the question U.S. defense and intelligence leaders asked at the recent GEOINT 2017 Symposium.
It is surreal to see our geospatial intelligence (GEOINT) community embrace and promote automation and data science to realize the potential of anticipatory intelligence. We have been working toward this vision for more than a decade. In fact, in the mid-2000s, as trained intelligence analysts, it was the promise of bringing technological innovation to the hardest problems that initially attracted us to working for SPADAC, an analytics company that paired geospatial data and other information to predict future events primarily for defense and intelligence customers. Founded in 2002, SPADAC’s goal was to bring advanced geospatial analytics from federal labs into the commercial environment. SPADAC was acquired by GeoEye in 2010, which catalyzed our analytics capabilities. This was followed by DigitalGlobe’s acquisition of GeoEye.
While the name of our organization changed, we continued our work—applying predictive analytics and machine learning to diverse sources of geospatial data to “narrow the search space” and focus scarce resources. We think of this as being able to answer “show me where” questions. One of the primary tools we use is Signature Analyst. This patented tool and methodology enables analysts to examine thousands of geospatial data layers to discover spatial relationships, patterns and preferences associated with different forms of human activity. This anticipatory intelligence tool can be leveraged in a variety of ways to support commercial, military and intelligence requirements. Over the past decade, it has supported counter-terrorism, counter-proliferation, surveillance and reconnaissance optimization, and counter-drug trafficking, as well as identifying areas most susceptible to radicalization. For example, we helped global development organizations (GDOs) determine the best locations to drill water wells and build schools to avert radicalization.
But we did not develop this technology overnight or without help.
In the mid-2000s, SPADAC was selected to participate in the Small Business Innovation Research(SBIR) program, which encourages small companies to perform innovative federal research that has the potential for commercialization. Our work on geospatial predictive modeling advanced to Phase III status and it was validated when we were awarded three patents for our approach (US 7346597 B2, US 7571146 B2 and US 7801842 B2). We combined these technologies to create Signature Analyst, our analytic modeling tool that gives geospatial analysts an unbiased statistical approach to examining events within an area of interest, discover relationships between events and factors, and predict where similar events might occur.
In this image from Signature Analyst, the model is highlighting areas in Jakarta, Indonesia most likely to experience extremist violence.
Over the past decade, we took an idea, made it tangible and proved it can work. We have since been applying and evolving Signature Analyst for a variety of large-scale geospatial missions for some of the most demanding customers across the U.S. Defense and Intelligence community. This continued investment in Signature Analyst and its value to our customers has led to three customers awarding DigitalGlobe SBIR Phase III sole-source contract awards over the past 18 months including the DIA contract we announced in June of 2016. We are developing automation and crowdsourcing capabilities to collect and enrich geospatial data at scale to allow our customers to enable Signature Analyst predictive modeling anywhere in the world.
Bigger data and cloud computing abilities enable us to evolve the architecture of Signature Analyst to facilitate analytics at scale. For this reason, this summer we will release updates to Signature Analyst Desktop and begin beta testing our cloud-enabled offering, Signature Analyst Server. We believe Signature Analyst Server will be a game-changer that allows our customers to answer new types of “show me where” questions and share insights with a broader audience of intelligence analysts.
This Signature Analyst output highlights areas in Soyanpango and Ilopango, El Salvador most likely (top 10 percent) to experience gang violence.
Signature Analyst will expedite anticipatory intelligence at scale because it is part of the DigitalGlobe—an ecosystem that provides access to global satellite imagery, Human Landscape foundation data, the crowdsourcing powers of GeoHive and Tomnod, and machine learning algorithms like those integrated into DeepCore. We have the compute power to process these datasets at scale through our Geospatial Big Data platform (GBDX) and the ability to build and deploy Signature Analyst Desktop and Server in AWS. Our decade of experience in the tradecraft of delivering anticipatory intelligence enables us to know the right questions to ask to get the desired information. We continue to develop novel approaches to realize our vision with further patents (US 9589210 B1, US 20170061625 A1).
DigitalGlobe has been building up to this moment, and we are now positioned to support the types of missions at scale that Lt. Gen. Stewart and other leaders asked for at GEOINT. We are excited to help our customers realize the future of anticipatory intelligence by harnessing commercial innovation. At GEOINT 2017, NGA Director Robert Cardillo nicely summed up why anticipatory intelligence is important:
“We know more about our planet—and any given emerging threat—than at any time in history. And we’ll be able to anticipate opportunities and threats as we provide true decision advantage over our adversaries. As we do so, we will steal space and time from those who seek to harm our security and provide it to those charged with protecting our freedoms.”