oldiers from the 94th Army Air and Missile Defense Command, track an unmanned aerial systems (UAS) threat during a scenario as part of Black Dart 18 on Muscatatuck Urban Training Center, Indiana, Sept. 17

oldiers from the 94th Army Air and Missile Defense Command, track an unmanned aerial systems (UAS) threat during a scenario as part of Black Dart 18 on Muscatatuck Urban Training Center, Indiana, Sept. 17 (U.S. Photo by Sgt. Sarvesh Regmi)

Learn to Use Data or Risk Dying in Battle, New Army Project Teaches

Project Ridgway pushes soldiers to use—and even create—the artificial-intelligence tools that will confer military advantage.

FORT BRAGG, N.C. — Soldiers aren’t rushing to use the artificial-intelligence tools that Pentagon leaders are rushing to develop, so Army leaders at Fort Bragg are launching a training program meant to convince commanders that trusting data, algorithms, and AI will keep them alive in battle. 

Called “Project Ridgway,” the nascent program aims to help soldiers and commanders better understand how the data they are holding and producing can yield important advantages on the battlefield. Col. Dan Kearney, the XVIII Airborne Corps plans officer who developed the program, told Defense One in an exclusive interview that contemporary military training just does not adequately address the importance of digital data in future combat. Soldiers and operators, Kearney said, are accustomed to putting data into PowerPoint slides, but not helping machine-learning tools to output insights. That could put the force at a big disadvantage in fights against high-tech competitors whose soldiers know how to put AI to work.

“We have to put the force in a position so that when artificial intelligence efforts are thrust upon us, we are in a position to go ahead and employ them immediately,” he said. “You want leaders to trust that the algorithm is going to go ahead and extrapolate from the right data sets to go ahead and come out with this recommendation. And if commanders don’t trust it, they won’t use it, and they’ll second-guess it, and it will slow down the kill chain.”

Kearney came to Fort Bragg in May 2020 fresh from a fellowship at the Fletcher School of Law and Diplomacy, where he analyzed how nation states will use AI to secure strategic advantage. At Bragg, he reconnected with Lt. Gen. Michael Kurilla, the commanding general of XVIII Airborne Corps. Kurilla told him he wanted to “operationalize innovation and build this AI-enabled corps. How do we get there?” recalled Kearney. That’s what Project Ridgway aims to do.

If it’s successful, the project could provide a blueprint for how the rest of the Army—and possibly other services—prepares soldiers and operators to use and build AI tools. Kearney said this starts with removing the mystique around AI and breaking it down into its core components: data, computing power, and algorithms. 

While popular culture tends to focus on writing algorithms, the first two items on that list are actually much more important.

For companies like Facebook, Google, and Amazon, the reasons even low-level artificial intelligence is suddenly much more useful today than it was just a few years ago have little to do with advances in creating new statistical methodologies. It stems from two concurrent phenomena: The relatively recent availability of enterprise cloud storage means that programs aren’t limited by hardware. Even more important, far more data is available. But that data must be collected, saved, and stored in a way that allows machine learning and AI programs to use it. 

Big tech companies understand that as a core aspect of doing business in the 21st century. Every search, friend interaction, or “like,” is essentially an act of data structuring, one that users take on voluntarily in exchange for free access to digital services. 

But military personnel are accustomed to thinking of the data under their purview as having one audience—their superior—and one form—a PowerPoint slide.

“They don’t really appreciate or understand data as a strategic asset,” Kearney said. “They save items on their desktop, save it on the portal; they save it on a shared drive. They save it and they store it based on the way that they grew up. Unfortunately, it’s incredibly difficult to aggregate all that data when it’s stove-piped in all those different silos.”

Changing that mindset is one of the first things the project will seek to change. 

The first line of effort seeks to “create a culture of innovation that respects data as a strategic asset,” Kearney said. That will happen through large-scale basic, voluntary training on the role that data plays in artificial intelligence and other emerging technologies. That part of the project will target both executive leadership and Army personnel who save and manage data—the people making those PowerPoint slides—to teach them a different way via basic courses, about two hours per week. It will launch on a large scale across the XVIII Airborne Corps next spring, but already some 250 people have finished a four-hour basic course.

The second line of effort will apply to those soldiers who want to develop more skills to manage data and even build apps. That starts with four months of training through the software coding bootcamp Galvanize, followed by six months of training through Joint Special Operations Command’s Gap program, learning to build apps that respond to real-world problems. There’s also training in applied data science through the education company Coursera.

Soldiers at Ft. Bragg who are part of the initial rollout already are building apps that  serve their needs. Chief Warrant Officer Brian Masters developed a program to track vaccination rates on base in real-time. Spc. Morian Senador built a data-visualization tool for commanders to better understand objects in their inventory. Both are full-time soldiers with other duties, but were able to quickly build low-code apps as part of the JSOC fellowship and their involvement in Project Ridgway.

The final two lines of effort will build off the first two. One will focus on data governance: Why does one service branch, or even one part of a service, have different names and standards for structuring data, and what can be done to overcome that? The other will focus on cloud computing, data storage technologies, and other elements of data infrastructure. 

Achieving a level of real data literacy across the force is a far-reaching goal, Kearney said. 

“You look at the definition, that’s probably something that’s out of reach for the workforce unless we start inculcating that into all of our professional military education across all of the [military occupational specialties] and branches, and that’s not something we can do in the XVIII Airborne Corps,” he said.

But, he said, these first steps could begin to put the Army on a path to better prepare soldiers for the sorts of challenges they’ll face in a world dominated by new technologies.