NATIONAL HARBOR, Md.—Dozens of onlookers gathered on a pier overlooking the Potomac River, shivering in the cold wind to see a small, 24-foot long vessel speed around without a captain.
It was Corsair, the smallest offering from robot-boat startup company Saronic, demonstrating how well it could maneuver in choppy waters with other craft passing by. And while it drew the largest crowd of the day, it wasn’t the only autonomous or semi-crewed vessel docked just outside of the Sea-Air-Space conference this week: Maritime Robotics, Textron Systems, and Ghostworks’ optionally-crewed Minerva all had a mooring. Project Perfect Storm was also there.
Paired with high-profile uncrewed undersea vehicle announcements from Anduril, HII and others, the expanded waterside showcase this year highlighted the Navy’s growing calls for more robots in maritime operations—both in the air and at sea.
It’s been a long time coming for Rep. Trent Kelly, R-Miss., who chairs the House Armed Services Subcommittee on Seapower and Projection Forces.
“The Navy has talked about it, talked about it, talked about it, but they’ve taken no actions to actually do this,” Kelly said of fully embracing the use of uncrewed maritime systems. “And I think we, as industry, and we as Congress, need to think ‘there are some platforms that can be both—can be either [crewed or uncrewed]—depending on the mission. So we absolutely have to focus there.”
Last year, Adm. Lisa Franchetti, then the chief of naval operations, outlined a 15-year path for regularly incorporating uncrewed systems as standard in maritime operations. That sentiment was bolstered by the CNO’s 2024 Navigation Plan, which highlights robotic and autonomous systems as key to increasing naval fleet capabilities. And Navy leaders are looking ahead to incorporating uncrewed tankers—the MQ-25—with aircraft carriers later this year after long delays with the program.
Yet there’s an appetite for more, particularly in the Indo-Pacific region.
During his testimony to the House Armed Services Committee onWednesday, Adm. Samuel Paparo, the head of U.S. Indo-Pacific Command, told lawmakers the command requires “sustained investment” in “autonomous and artificial intelligence-driven systems” and other technologies “to maintain credible deterrence.”
In written testimony, he said “expanded acquisition of additional autonomous aerial systems, AI-driven undersea vehicles, and enabling technologies for full deployment will provide critical capability in the Indo-Pacific.”
The Navy is also testing two torpedo-tube launch and recovery underwater vehicle systems—HII’s Remus 600 called “Yellow Moray” and L3Harris’ “Rat Trap” or Iver4—with the United Kingdom, as part of the trilateral AUKUS agreement.
“We’ve tested them on both Los Angeles– and Virginia-class submarines. We currently have [USS] Delaware deployed with the Yellow Moray system. And the [Royal Navy’s] Astute-class, over the last five or six months has tested the UK version or variant of those systems as well,” Vice Adm. Rob Gaucher, the commander of Naval Submarine Forces, the U.S. Atlantic Fleet, and Allied Submarine Fleet, said during a panel discussion.
But the Navy’s hybrid fleet isn’t yet a reality.
The Navy isn’t new to the concept of uncrewed or autonomous systems. The idea started to take shape with the Navy’s 2021 Unmanned Campaign Framework, which outlined how the service wanted to use the systems.
The document was an “entry salvo on how it might work, but it never really matured to the point of execution,” John Mustin, the president of Saildrone, which makes autonomous vehicles for oceanographic and bathymetric surveys and defense and maritime domain awareness applications, told Defense One on the show floor.
In the years since, there’s been a greater acceptance of the reality that a naval fleet must have uncrewed systems.
“We will continue to build ships, submarines, and airplanes, no question. But we also know that we’re going to incorporate unmanned systems, some of which are autonomous—not all are, but many of them will be,” said Mustin, who was previously the vice commander of U.S. Fleet Forces Command and commander of the Navy Reserve.
“It could be as much as a third to a half of the total number of platforms. And the benefit of autonomy is they don’t need to be big and you don’t need a lot of people on it,” he said.
And fewer people means lower costs for training, and potential lives saved.
“When I look at a $2 billion destroyer that has 350 people on it, and I think of the overhead associated with the training that is required to man—from an O-5 commanding officer to an E-1 line handler,” Mustin said. “The benefit here is we provide all of the operations and maintenance. So you just get the benefit of the capability.”
So if the benefits of autonomous systems are clear, what’s the hold up? Charles Fralick, Leidos’ chief technology officer, said the Navy’s reticence is not because the technology isn’t there, but because there’s uncertainty about how to deploy them.
“A small underwater vehicle or small surface vehicle [or] small air vehicle—they are wonderful pieces of hardware today capable of doing many really interesting things, as you’ve seen in Ukraine. But the issue is, you have to get them to the point that they need to be used, and you need to be able to support them where they are deployed,” he said.
Smaller uncrewed systems have limited ranges and power, which stunts their ability to maintain communications at great distances.
“That means recharging underwater vehicles, refueling surface vehicles and air vehicles, and having the communications infrastructure. And so that’s the piece that the Navy has been starting to come to grips with,” Fralick said.
Larger platforms are easier to deploy, can handle military communications, and have long ranges and high reliability, he said, but come with their own challenges.
Rear Adm. William Daly, the Navy’s director of surface warfare, recently questioned the service’s need for large autonomous ships, arguing instead for a medium-sized vessel that can be versatile and produced quickly. Lawmakers also doubt the need for large USVs.
Medium-sized USVs can do just about everything the large ones can—like contested logistics and carry weapons—except really heavy lifting. But there’s still that deployment problem, Fralick argued.
“[It’s] not just getting the things from point A to point B, where they can be used, but it’s also supporting them once they get out there,” he said.
That is, a USV that’s hundreds of feet long will likely require some of the same specialized maintenance that causes backlogs and delays with warships, whereas smaller platforms can be made faster and cheaper.
“I think it’s the Navy optimizing the balance between production efficiency, cost, and capability,” he said. “We will eventually get to large USVs in Navy service. But there’s no ultra-compelling demand signal right now because we can satisfy most of the needs with the MUSVs” and smaller.
Challenging tides ahead
There are challenges besides policy to thoroughly weaving uncrewed systems into the fleet from the Defense Department’s complicated acquisition system, which the Trump administration has vowed to reform. Namely, limited manufacturing capabilities and contested communications.
“You still want command and control folks ashore or afloat to be able to reach out and touch these platforms and reprogram them in extremis,” such as if they’re surrounded by dozens of ships, Fralick said. “To have reliable communications no matter what scenario we’re in is very important, and we’re making good progress in that area. Five years ago, 10 years ago, that was a real issue.”
Another tech challenge is how well the systems can see or perceive their surroundings.
“Most builders of autonomous surface platforms use radar as their primary sensor,” he said. “Radar is good in many situations; if you have calm seas or you’re not in a wartime footing where you can radiate from the radar, they’re pretty good.”
Fralick believes in having layered solutions—optical, LiDAR, and also AI—to produce “a fused perception picture.”
“We have been a little slow in making progress in that area in the community,” he said. “But that’s got to grow at a more rapid pace to ensure that we have confidence in our ability to operate in contested environments, in all weather conditions, because the worst possible outcome with an autonomous platform is that it collides with a pleasure boat and kills people.”
A water taxi with passengers ferried by as Corsair effortlessly made shapes, trailed by a manned safety craft. It was impressive to see, but also a reminder of why there’s been hesitation on full-scale adoption.
“We’ve got to make sure that doesn’t happen. The autonomy is good enough…It’s going to take something really unusual, like a deliberate attempt, collision by another vessel, for us to hit a platform. It’s very safe, but it’s only as good as the perception systems,” Fralick said.
There’s also a need for systems to detect and track other vessels at a distance. Watching Corsair cut through the wind-made waves on the Potomac became more difficult the further away it got from the pier, until it was identifiable mostly by its foamy wake.
“Even with the naked eye, if you see a vessel several miles away, unless it’s very large, you have a hard time telling what it is until it gets much closer. We need the ability to do that more quickly and with confidence on board our autonomous platforms,” he said.
Of course, there are different levels of autonomy—from the kind that presses the brakes to avoid a crash to the kind that does everything on its own, even if there’s the ability for human override.
“They’re not quite at the point where they’re going to allow an autonomous platform to do everything, culminating in the launch of the missile. We’re not quite there with the trust level yet,” Fralick said. “That’s coming, but that requires a lot of testing to ensure the safety of civilians and other platforms that aren’t adversaries. That needs extra care and attention, and it’s going to take some time until the Navy is comfortable with that.”
Jesus Christ is King
