DARPA Chief Outlines Unmanned Military Tech Advancements

The Defense Advanced Research Projects Agency (DARPA) is advancing unmanned military technology to “radically transform” today’s battlefield and contribute to future military programs, DARPA director Stephen Winchell said last week at the 2025 Global Aerospace Summit in Washington, D.C.

Winchell pointed to successes like the No Manning Required Ship (NOMARS), DARPA’s autonomous ship program, which produced the USX-1 Defiant, an unmanned surface vessel capable of operating at sea for extended periods of time without the need for human operators onboard.

“The real focus of it was not on the autonomy stack. There’s some of that, but it was mainly focused on building a really easily manufacturable robot that could be maintained all around the country [and] built in tier three shipyards,” Winchell said.

The ship is still undergoing testing, but could become a key component of the Navy’s push to increase the size of its fleet by 2030 with manned and unmanned vessels.

Former Secretary of the Navy Carlos Del Toro said last year that unmanned vessels could assist worldwide maritime operations and interdiction of illegal drugs and arms.

“We need to fully explore how we would develop the concept of operations for large, unmanned surface vessels. But I think investments in medium unmanned surface vessels make all the sense in the world,” Del Toro said at the time.

Failure Is Part of DARPA’s Innovation Strategy

Winchell said the more than 280 programs he oversees operate on different schedules, and the agency’s nearly $5 billion budget must be “balanced in risk and time and space across domains.”

However, Winchell noted that failure is an important part of DARPA’s operations. Approximately 70% of DARPA’s programs don’t meet their metrics, but the rubble of failed programs is often repurposed for other uses.

Winchell pointed to the Liberty Lifter, DARPA’s unsuccessful attempt to create a cargo seaplane capable of operating without runways. The program was canceled in June after officials determined it would not be cost-effective to build the aircraft.

“We failed, I think, for the right reasons, which are bumping into the limits of physics, bumping into limits of material science and engineering,” Winchell said. He added that “realizing that you’ve got to pivot the program and take the good things you learn from the sensor systems, the autonomy, into somewhere else,” meant the work done by the Liberty Lifter team was not in vain.

He said that compassion, honesty and a strong workplace culture encourage employees to admit when a program can’t get off the ground, freeing up resources for other ongoing projects.

“Ideally, people cancel their own programs,” Winchell said. “They say, ‘Listen, I can’t do it. I want you to take this money back and do something better with it.’ Ideally, also they’re doing multiple programs so their eggs are not in one basket.”