Army looking to hide bases, speed up anti-drone capabilities amid modernization drive, Under Secretary says
The service also sees AI as key, but warns that data management is a major hurdle.
In America’s 20 years of war in Afghanistan, the Army had one key mission: find the enemy.
On the battlefields of the future, they may add another, competing mission: hide.
The ability of America’s enemies to easily observe, track, and strike bases is just one of many factors the Army is considering as it pursues modernization, with an eye on everything from drones to artificial intelligence, Army Under Secretary Gabe Camarillo said in an exclusive interview with Defense One.
The war in Ukraine has shown that an easily found headquarters is immediately vulnerable to attack. Using a combination of intelligence and long-range weapons, Ukraine has struck numerous Russian command posts, resulting in the deaths of at least two Russian generals.
“The Army is recognizing that reducing signatures and the ability to hide in plain sight is going to be a very important characteristic moving forward,” said Camarillo.
To handle the problem, Camarillo said he and Army Vice Chief of Staff and Chief of Staff nominee Gen. Randy George are working together to re-evaluate networking and communications systems used in headquarters.
The many communication antennas that sprouted from headquarters in Afghanistan are likely a thing of the past, Camarillo said. The service also wants to field equipment that is less visible on the electro-magnetic spectrum, and therefore harder to target. New equipment will have to “comply with the realities of future operating environments,” Camarillo said. “That's very much feeding into the requirements process today.”
A future adversary is also far more likely to track U.S. soldiers with drones, sending in loitering munitions for quick attacks or coordinating dozens of pinpoint-accurate artillery strikes.
Countering this threat will require the Army to not only field more anti-drone capabilities, but also rethink how it produces weapons, Camarillo said.
While most Army development programs create tools to answer unchanging specifications, the varying nature of drones means a single enduring weapons platform may not be enough to get the job done.
“I think we've got to come into a situation where it's an ongoing cycle, just like we do with software, where it's under continuous development,” said Camarillo. “It's going to challenge us to find new ways of doing business.”
Directed energy weapons, including microwave technologies that have proven effective in taking out drones, are among the Army’s efforts to tackle the problem, Camarillo said. The Army’ Joint Counter-small Unmanned Aircraft Systems Office, founded in 2020, is also simultaneously exploring gun and missile technologies for defeating drone attacks.
Not everything is changing though—despite increasing buzz about artificial intelligence, the Army has some long-known barriers that bring visions of an AI-guided future back down to Earth.
For one, the Army’s data is less consistently formatted and more difficult to aggregate than the reams of high quality data that have powered fast-moving rises in AI capabilities.
Creating large enough data pools for high quality artificial intelligence is a “very, very difficult task," Camarillo said, noting that data from Army intelligence and weapons systems are managed “very differently” by the separate Army units responsible for them.
Gaps in the workforce are another hurdle that would remain even if the Army solved its data problem.
The Army must “upskill its workforces to understand how these algorithms are developed, how they can be applied, what datasets are required,” Camarillo said, noting an Army Futures Command partnership with Carnegie Mellon to offer AI-related training to soldiers.
Slow Defense Department networks may be another limitation—a complaint that is perhaps already familiar to those who have publicly complained of waiting as much as an hour to log into computers, or of home pages that crashed upon opening Excel.
“The ability of our tactical networks to pass data at the volumes that would be required,” will be a factor in adopting artificial intelligence, Camarillo said. “The pace of our expanded use is really driven by our ability to unlock data.”
The artificial intelligence revolution, it seems, is likely not coming in the form of killer robots, but killer spreadsheets.
Listing out high-value potential applications for AI, Camarillo noted intelligence, predictive logistics, and audits, a role Camarillo is particularly eager for, given his role as overseer of the Army budget.
“As we continue to expand our use, it will likely have unexpected applications,” Camarillo said