Artificial intelligence and other automation technologies are essential for the Department of Defense (DoD) to maintain its technological edge over adversaries like China – but increasing the use of both of those will require more funding and leveraging of industry capabilities, witnesses said during a House Armed Services Committee hearing on Monday.
At a committee field hearing held at the University of California-Santa Cruz Silicon Valley campus in Santa Clara, Calif., on Sept. 16, members listened to witnesses explain how DoD can overcome its “struggle” to rapidly develop, scale and innovate its warfighting capabilities.
“For years, we’ve been hearing complaints from industry about the glacial pace of acquisitions and from small innovators that lack the capital and support necessary to bridge the valley of death. This is immensely frustrating to us,” Rep. Mike Rodgers, R-Ala., who chairs the committee, said during an opening statement.
“Earlier this year, we held a similar hearing with senior DoD officials who insisted they had all the authorities they needed and were doing everything they can to expedite innovation. We’re here today because we respectfully disagree. We think more can be done and should be done,” he continued.
Witnesses focused on drone technology and drone-fighting capabilities, explaining that AI systems allow for a single operator to distribute intent to multiple robots – including those that are operating in different environments including air, water, and on the ground.
Mark Valentine, president of global government business at Skydio, said that drones are a “game changer” and “transforming warfare.”
“Drones are smaller, they’re smarter, they’re attritable, and often they are as important to ground personnel as the rifles that they carry – in short, these drones are no longer nice to have, they’re absolutely essential on the modern battlefield,” said Valentine. “On the battlefield in Ukraine, troops rarely maneuver or even fire a shot unless they have some small [intelligence, surveillance, and reconnaissance (ISR)] drone in the air.”
Other technologies discussed at the hearing include AI pilots that enable self-driving aircraft. Brandon Tseng, the co-founder and president of Shield AI, said that while his company has made progress, automation tech receives funding at “levels that lack credibility or seriousness.”
“This technology enables unmanned systems to execute missions without GPS communications or remote pilots. It also enables the concept of swarming, which enables a single person to command hundreds of thousands of drones effectively on battlefield with Russia, China and Iran,” said Tseng. “Self-driving autonomy has become the single most important technology since stealth and GPS-guided munitions.”
By shifting about 25 percent of funding from a requirements-based acquisition system to a problem-based acquisition system, better DoD access to new technologies like AI pilot systems can be realized, Tseng said.
“The DoD acquisition process is not built to solve problems; it is built to fulfill requirements which takes anywhere from three to 20 years to validate and budget,” he said. Using problem-based acquisitions would provide more spending flexibility and investment in innovation, he continued.
Other witnesses from Palantir, Applied Intuition, and Saildrone, Inc., echoed similar calls to expand and strengthen collaboration with the commercial industry to proliferate new technologies.
Recommendations included developing a “thoughtful program design” to incorporate continuous development and integration into the DoD; expanding software acquisition; establishing dedicated bridge funds within service budgets; and providing combatant commands with five percent of the DoD’s budget to “enable strategic competition with services” to “beat back the worst instincts of the monopsony,” or a market where there is only on buyer.
