A new Pentagon program is pushing drone makers to continuously improve their systems based on troops’ feedback, hoping to spur innovation that moves as quickly as the war in Ukraine.
Launched on Monday, the Defense Innovation Office’s Project GI initiative aims to embed frontline insights into a perpetual loop of design, testing, and deployment. It’s a deliberate effort to mimic how the Ukrainian military has out-innovated Russian forces by rapidly fielding and iterating drone technology under fire.
The chaotic pace of Russian electronic warfare tactics necessitates that, a factor that is likely to exist as a backdrop to more future conflict, particularly in the Asian-Pacific. It will be open for submissions on a rolling basis through December 31, (which is also a change from the traditional narrow submissions window for competitions.)
The Pentagon has been exploring ways to more rapidly acquire drones since the 2017 operations against ISIS in Syria and Iraq. While it has made some progress—particularly in the special operations community—those efforts largely reflect streamlined bureaucracy rather than new mechanisms for real-time adaptation and innovation, Trent Emeneker, DIU’s Blue UAS lead, said in an interview.
Rapid equipping may be one of the Pentagon’s stated goals, but “I think the reality is we see zero evidence of that happening,” Emeneker said. Even so-called “rushed efforts oten take years to make it through the requirements writing process. ““Then they’ll need to get budgetary approval. That’ll be another two to three years,” he said. “And that’s the definition of fast within DoD today: five years to even start prototyping—not delivery, but prototype.”
Adversaries large and small—and increasingly aligned with one another—don’t face the same hurdles. Ukraine shows how modern warfare has become a contest to deploy attack drones adapted to fast-evolving enemy countermeasures.
Emeneker, who has spent considerable time communicating with Ukrainian forces and other innovators, said arms makers need far closer and more constant contact—not with contracting officers but with operators on the front lines.
One of DIU’s Blue UAS companies is already incorporating feedback from Ukraine’s soldiers into its drone designs. “They have roving teams of engineers that go to their partner units constantly and get feedback within a kilometer or two of the front lines. They live with the guys,” he said.
One of the companies “living” alongside Ukrainians is Shield AI. Co-founder and president Brandon Tseng told Defense One that the company has teams working in Ukraine around the clock, which helped Ukraine adapt last August to intense Russian GPS-jamming that hurt their ability to target Russian missiles. Speaking on podcasts before the 2025 GLOBSEC Forum in Prague, Tseng said Shield AI’s Ukraine team “communicated the problem back to the team in the United States. Software engineers got to work, worked throughout the entire night, tested it the next day at a range in the United States, and then sent it to our forward-deployed engineers who implemented it” with 24 hours, enabling the destruction of Russian missile systems.
“The Russians will start to do things where we have to be very, very proximal to the problem, so that we can implement these quick changes. That’s the speed of warfare these days. If you want to be relevant again, you have to be right next to the problem,” he said.
By normalizing that sort of deep, continuous involvement, Emeneker believes, the new program will dramatically reduce the time it takes to build or modify drone prototypes—from years to a couple of months for hardware, and far less for software updates.
“We want to start trying to solve the problem in 24 hours,” he said.
But Emeneker added that producing drones in sufficient quantity depends on reducing U.S. dependence on Chinese control over key components in the drone supply chain—items like magnets for actuators, lenses and other critical parts.
Even if the United States were to fully commit to urgently reshoring drone parts manufacturing, Emeneker estimated it would take more than a year before production could scale to meet potential demand. “You start to see incremental delivery improvements probably in six months. As we get to the nine- to 12-month mark, we start to see measurable improvements,” he said. “As we get to 15 months, we really start that hockey-stick curve of exponential deliveries.”
Read the full article here