Enable breadcrumbs token at /includes/pageheader.html.twig

JADC2 Begins With Intelligence

Data centricity may change everything.

The U.S. military’s concept for Joint All-Domain Command and Control (JADC2) begins with intelligence data, and data-centric operations will require profound changes, according to a panel of experts.

Those experts were speaking on the final day of the May 25-27 AFCEA Spring Intelligence Symposium. The panel included Stuart Whitehead, deputy director, cyber and command, control, communications and computers integration, Joint Staff J-6; Preston Dunlap, chief architect, Department of the Air and Space Force; Col. Andre' Abadie, USA, JADC2/Project Convergence lead; and Van Hendrey, program manager, Machine-assisted Analytic Rapid-repository System (MARS) within the Defense Intelligence Agency. Lewis Shepherd, who serves on the AFCEA Intelligence Committee and teaches big data in the Graduate Department of Information Sciences and Technology at George Mason University, moderated the panel.  

Whitehead said in his opening remarks that JADC2 really begins with intelligence, including signals intelligence and human intelligence, and the ability to quickly and effectively process data. “At the heart of JADC2 is our ability to operate faster than our opponent, to make decisions faster and to make better quality decisions along the way,” Whitehead said.

The emphasis on data requires a greater emphasis on interoperability as well. “In order to operate at that appropriate level of speed, that requires a lot of machine-to-machine transactions. That means that we have to have probably more technical rigor in how we approach interoperability than we have in the past,” Whitehead added.

He made the case that the focus on data is a bona fide paradigm shift for the military. “Literally we are moving from net centricity to data centricity. Those aren’t just words. That’s a significant change in how we think about data, how we move it, how we store it, how we secure it, and where a lot of the transactions and processes occur,” he explained.

For one thing, JADC2 will change the military’s relationship with intelligence. “We’re not going to have the ability necessarily to always have reachback. We sure would like that, and we’ll do everything to ensure that we can, but we are recognizing, particularly for major combat operations, our ability to store and process information, stream information, compute, will be also at echelon. That goes also for intelligence. If we think about the ability to sense and act in an isolated fashion, that’s part of the way forward.”

Also, data-centric operations will change the way military officials think about budget planning. “Because of data centricity, it’s really challenging the way we think institutionally. It’s unlikely that our current way of doing [budget planning] is going to be able to meet the challenge. We’ll have to have a more agile way of doing funding, particularly in a software-intensive environment,” Whitehead observed. “Those involved with software know you’re never done. So, we’re continually changing, and JADC2 as much as anything else, is a continuous journey and modernization activity that is unlikely to stop.”

During the question-and-answer session, Whitehead noted that there is no Defense Department-level budget for JADC2, but he expressed confidence the concept will not fail as other ambitious military programs have in the past. “As somebody may not know, there is no JADC2 budget. I have like zero dollars. I have a great argument, but I don’t have any money,” he said. “It means we have to work together and work smart.”

He suggested the department can save money by eliminating duplication, consolidating networks and reducing point-to-point interfaces and then apply those saved dollars to JADC2. Still, sometimes to save money, the department must first spend money. “It doesn’t mean that at the Joint Staff level or the DoD level we don’t need money. What we would be looking for is the ability to provide incentive funds. It takes money sometimes to close things down. If we want to accelerate, some resources will have to be acquired to do that work. We’re going to see a combination of all of that moving forward.”

One reason past programs failed was that the Defense Department and military services did not agree on the goals or how to get there.

It’s different with JADC2, according to Whitehead, who pointed out that the Defense Department leadership, the military services and the various cross-functional teams, working groups, combatant commanders and international allies are all working together. They also are asking the toughest questions first, such as how to share intelligence data with coalition partners, how to perform in a Zero Trust environment and how to implement multilevel cybersecurity.

“We’re building in that partnership from day one. These really tough questions that largely we’ve dodged for years, we’re hitting them head on. If we can solve those, that means a lot of the other things will follow,” Whitehead said.

For the Army, the JADC2 concept also could change the way the service establishes requirements for some technologies, such as artificial intelligence, Col. Abadie suggested. Traditionally, when service officials wrote requirements documents, they already knew how many systems they wanted to purchase.

But that is not necessarily the case with artificial intelligence. “When we talk about artificial intelligence specifically, we’re going to have to recognize that there will be a break point. There will be an echelon where these things will perform optimally because there will be dependencies,” Col. Abadie explained. “Therefore, it’s one thing to get it to work. It’s another thing to experiment with enough vigor at enough echelons to find where that optimal point is.”

That discussion of how far technologies will scale presents challenges for the Army. “As we’re looking at JADC2 for the Army, scale becomes a challenge—just the number of devices. I think the solution going forward is really going to be that ability to compute at the edge and ensure that whatever data you’re passing is the minimum amount necessary to provide relevant information to an overall common operational picture.”

The colonel told the symposium audience about a concept known as an artificial intelligence common operational picture, or AI COP. “What it means is that in the background, let the algorithms continue to build something. It’s continuing to build that picture because … everybody has different priorities and intelligence requirements.”

Advances in AI technologies also are shifting the Army’s thinking regarding its “sensor-to-shooter” concept. “By leveraging these artificial intelligence technologies, we’re going from all sensors to the best shooters to the right C2 node in order to get that convergence done,” Col. Abadie stated. “We need to expand our thinking. We always say sensor to shooter. It’s quite trite now. What if we said sensor to shooter to sensor? What if we asked ourselves to try to figure out how to do that at machine speed?”

In a Twitter direct message exchange following the panel, Shepherd described the sensor-to-shooter-sensor concept as a “whole new paradigm of what order of battle means in the fight.”

Meanwhile, the Air Force and Space Force are approaching JADC2 and data centricity with a lot of software-defined capabilities, according to Dunlap. “We’re software defining as much as we can. We’re building new software-defined data networks,” Dunlap said.  

In fact, just a week ago the Air Force and Space Force concluded the first-ever data and architecture infrastructure summit, a two-week event that brought together the two services, combatant commanders and “data owners of our big data platforms” to “evaluate different courses of action and “achieve very rapid success in integrating data in a way that makes a significant operational difference” and not just “Band-Aiding and duct taping,” Dunlap reported.

He added that next month, the leadership could approve “significant programmatic, budgetary and operational decisions on data.”

Hendrey updated the audience on the MARS program. Among other benefits, the system combines data from a variety of sources and presents it in a way that is familiar and easy for most users to understand.

“We are building a modern data environment consistent with commercial best practices to curate the data that exists in our legacy system today, transform it, make it ready for machine learning, and make it accessible and discoverable across the intelligence community as well as other services and users through JADC2,” she said. “For us, JADC2 is really exciting because while we are providing this data in a more accessible way, JADC2 really provides that connective tissue for us to then connect with the data that other services and other service systems are bringing into the fight.”

Shepherd, the panel moderator, noted that China has claimed to be launching a 13,000-satellite constellation as a competitor to the SpaceX Starlink constellation and asked whether MARS is prepared to handle the space order of battle during the current competition with China or during a potential war.

“We already have records in the existing system today, the legacy system as well as the data that has been ingested into MARS, on space infrastructure, space order of battle. That exists today,” Hendrey responded. She noted that the Defense Intelligence Agency is working with the Space Force and U.S. Space Command, the military services and the other intelligence agencies to better prepare MARS for future conflicts.  

Shepherd told SIGNAL Magazine that, in his view, there has not been a “frank discussion yet about the vastly increased demands and requirements” for the Defense Intelligence Agency and MARS. “Its foundational military intelligence responsibilities for warfighting are already drastically altered and underlined by vast reams of data that, say, convergence brings,” Shepherd said.

“And MARS really needs to be even more ambitious and pathbreaking that it has been” and cannot be just be a next-generation Modernized Intelligence Database, which MARS is replacing. “You heard Van talk about the new approach she’s bringing to it. Very important,” Shepherd concluded.