Defense Media Network

DARPA AI Prototype to Help Determine if Observed Events are True Threats

Service members at U.S. Indo-Pacific Command headquarters in Hawaii recently tested a prototype DARPA system designed to help military analysts and planners determine if observed events – such as increased force movements, cyber intrusions, and civil unrest – are unconnected occurrences, or if they’re part of an adversary’s coordinated campaign to achieve strategic objectives in a geographic region.

Operational representatives from the command’s intelligence and operations divisions spent three days in December trying out DARPA’s COMPASS tool suite. COMPASS, which stands for Collection and Monitoring via Planning for Active Situational Scenarios, analyzes large streams of data to uncover competition campaigns, and displays results that represent the evidence and the analysis behind each hypothesis.

COMPASS seeks to leverage advanced AI and other technologies to help commanders make more effective decisions regarding a competitor’s complex, multi-layered competition activity. Competition refers to actions – both non-violent and violent – designed to achieve geopolitical goals without provoking full-blown armed conflict. The ultimate aim of the COMPASS program is to provide theater-level operations and planning staffs with robust analytics and decision-support tools that reduce ambiguity of adversarial actors and their objectives.

“We kicked the program off less than a year ago and pushed very hard to have something tangible for military users to evaluate,” said Fotis Barlos, program manager in DARPA’s Strategic Technology Office. “Indo-Pacific Command provided great support for this experiment, and their feedback and insights remain invaluable as we continue to develop the AI technologies and the user interface.”

COMPASS sets itself apart from existing digital aids, which mostly provide historical trends, leaving analysts to try to figure out if what they’re observing is normal activity or not.

“COMPASS tries to put all the incoming data into what we call ‘lines of effort’ and then automatically generates multiple hypotheses,” Barlos said. “The AI, in essence, tells operators, ‘Based on what I’m seeing, I believe these activities are part of sequence, an adversarial competition campaign’ and provides that information in a graphical form for the human to evaluate.”

“Regardless of the sophistication of the underlying algorithms, there will always be some uncertainty about the hypotheses an AI system can produce,” Barlos said. “This is an area that COMPASS addresses directly. “For this experiment, we told each team of analysts and operators they would experience the equivalent of 10 days’ worth of events from four major actors in the Indo-Pacific area of responsibility – a nation state, a terrorist actor, a background actor, and a proxy actor for the nation state,” Barlos said. “The participants’ task was to use COMPASS to determine each actor’s lines of effort and do it as early as possible in the simulated 10 days. The fewer days it took the better.”

Throughout the experiment, operators tested three versions of the prototype provided by the COMPASS performers: Raytheon BBN Technologies, Systems Technology Research, and Lockheed Martin Advanced Technology Laboratories.

The main benefit of the experiment was to acquaint U.S. Indo-Pacific Command operators and analysts with a new tool and a new process to conduct operations in the competition space with near peer adversaries.

“Right now, analysts read lots of reports and try to figure out what’s going on – and they’re very good at it,” Barlos said. “But every time you have to write something down and associate it with something you saw yesterday or the day before, it becomes a challenge. Now they have a tool that puts it all together, providing what we hope will be a faster and more useful means of addressing adversarial competition.”

Following the experiment, Air Force Col. Joseph Musacchia Jr., division chief of the U.S. Indo-Pacific Command directorate in charge of critical infrastructure protection, anti-terrorism, and mission assurance said, “COMPASS has potential to be an extremely useful tool in the world of mission assurance, and demonstrates that we can discover patterns of activity at the ‘speed of need’ and then act upon them.”

DARPA is working with the intelligence, operations, and plans and policy divisions of U.S. Indo-Pacific Command for further collaboration and to make prototype COMPASS tools available for additional testing by the Command using real-world data.