Inside the $9 billion Texas startup building a drone armada for the U.S. Navy
A former Navy SEAL turned tech investor, Mavrookas had recently launched the startup Saronic Technologies with a vision for shoring up America’s defenses: cheap, fast autonomous boats that could be commanded in swarms, armed with sensors and weapons, and manufactured at factory scale.But first, he and his team, veterans of SpaceX and Anduril, needed a prototype. So they bought a raft off Amazon and began rigging it with $30,000 in off-the-shelf cameras, sensors, and motors in their bare-bones Austin warehouse.A month later, Ukraine used its own makeshift drone boat to strike a multimillion-dollar Russian warship. Soon, lawmakers and naval officials were traveling to Austin to see the prototype. “Everybody was like, ‘Oh, drone boats are massively effective against larger naval vessels,’” Mavrookas says. Just 90 days after launching, Saronic signed its first Navy contract.Since then, Mavrookas’s vision has become a strategic imperative. For years, the U.S.’s long-term strategy has focused on the South China Sea, where Beijing’s potential plans to take Taiwan involve a flotilla of autonomous carriers and stealth unmanned surface vessels (USVs), aka drone boats. But after the U.S. and Israel launched their attacks on Iran, the Strait of Hormuz, where Iran’s cheap aerial drones and kamikaze skiffs have upended U.S. war plans and disrupted the global economy, the demand signal, as they call it, is blaring. “We’re seeing our thesis play out in real time,” Mavrookas says. “We need this as a country.” Indeed, in early June, two U.S. airmen went down with their helicopter near the coast of Oman; roughly two hours later, they were rescued by 5th Fleet’s drone-focused Task Force 59—using one of Saronic’s Corsair USVs. It was a Navy first, and possibly a world first. Industrial Revolution: Saronic acquired and revived the Gulf Craft shipyard in Louisiana last year to speed up its drone-boat building efforts. (Photo: Daymon Gardner)Mavrookas’s sense of urgency defines Saronic’s rise into a defense-tech powerhouse with 1,400 employees, $500 million in government contracts, $2.6 billion in total funding, and a valuation north of $9 billion. The company is using the cash, from investors like Andreessen Horowitz and Joe Lonsdale, to create autonomous boats and larger ships capable of carrying cargo, weapons, or flocks of more drones, along with the software for operating them.Saronic has several vessels in various stages of production, from the 24-foot Corsair to the 160-foot Marauder. Last year, it bought an ailing shipyard in Louisiana and built its first Marauder in nine months, making it possibly the fastest ship built from scratch in the United States since World War II. Saronic intends to build as many as 20 there next year, depending on demand.For Mavrookas, drone boat factories are the answer to a geopolitical math problem. Global power is, to a large degree, a function of sea power. China reportedly surpassed the U.S. in total number of naval ships about a decade ago. The country now accounts for 51% of global shipbuilding, while the U.S. turns out only a few dozen naval and commercial ships a year.Washington still leads in firepower, but in a prolonged conflict, the core issue isn’t just who has the most ships, it’s who can build them the fastest. “The risk isn’t just a fleet-size gap,” says Mavrookas. “It’s an industrial-endurance gap.”Faced with that imbalance, the Pentagon is increasingly betting on sea drones. The Navy has already allocated $3.7 billion for USVs, while the latest Pentagon budget request calls for billions more, and $65.8 billion for shipbuilding alone—the largest such request since 1962.Meanwhile, the Navy’s plans for a future “Golden Fleet” have a roughly equal mix of manned and unmanned vessels, including a new medium-size drone ship capable of hauling two payloads some 2,500 nautical miles—exactly what Saronic’s Marauder is designed to do. Navy officials envision around 30 such ships in the Indo-Pacific by 2030, along with thousands of small USVs.There’s little doubt that drones will transform naval warfare. But how quickly the U.S. Navy will adopt them and how they’ll actually be deployed are murkier questions. The answers could determine Saronic’s fate and much more.For more than a decade, the Navy has run quiet drone experiments and even deployed a handful of smaller vessels, including in the Persian Gulf. But no autonomous boat system has yet achieved enough “hands off” reliability to operate with weapons in a real-world environment. That puts the U.S. behind geopolitical rivals.“The time is now, and the Navy is late to the game already,” Senator Rick Scott said at a recent Senate hearing on maritime drones.A breakthrough came in December, when the Navy announced that it had awarded Saronic one of the department’s largest drone boat contracts to date: $392 million for the Corsair, a stealthy 24-foot boat with an estimated price tag of roughly $1 million that the startup makes in its Austin facility. Then–Navy Secretary John Phelan celebrated the milestone on X: “Prototype to production in under 12 months. This is now the standard.” (Phelan was fired in April, reportedly over clashes with Defense Secretary Pete Hegseth.)(Photo: Daymon Gardner)For Mavrookas, the contract’s dollar value isn’t as important as the signal it sends: that after years of playing with prototypes, the Navy is ready to move quickly. The message, Mavrookas says, is: “We want to reward companies that are investing their own capital to build the best capabilities for our warfighters and deliver them at speed and scale.”That production capacity starts with a simple ship design. Since Saronic’s vessels are uncrewed, they don’t need galleys, bathrooms, and sleeping quarters. The company is able to “strip the complexity out of the ship,” Mavrookas says. Its boats need only hold essentials such as engines, batteries, payloads, and AI systems that include cameras, Nvidia GPUs, radar, and other sensors.These systems are controlled by Saronic’s software platform, Echelon, which lets a human operator manage a mix of surface, subsurface, and aerial drones. Using natural language, users can define the objectives of a mission, simulate outcomes, and execute operations remotely; if they lose communications with a vessel, it can complete the mission on its own.As Saronic expands deeper into defense and soon commercial vessels (in addition to establishing operations in the U.K. and Australia, it has a partnership with offshore energy services company Hornbeck), it’s not hard to see a larger vision taking shape: Echelon serving as a hub for an ocean’s worth of drone data and a key part of the Navy’s command and control systems.Similar platforms are in development by half a dozen other drone firms, including Havoc AI, Saildrone, and Anduril, which are all jostling for Navy contracts. This is where the most profits tend to lie, says Joshua Tallis of the Center for Naval Analyses. If you’re a venture investor in this space, “you’re not going to make that money back selling aluminum speedboats to the U.S. Navy,” he says. “You want to be in the business of software or data streams.”But Mavrookas’s ambitions go beyond software. He wants to do for ships what SpaceX did for rockets—help revive and revolutionize an industry America once dominated. (The two share many investors.) To solve the shortage of skilled shipbuilders, he’s betting on what he calls a “cultural upgrade”: recruiting young engineers and upskilling autoworkers to build the next generation of vessels.(Photo: Daymon Gardner)He wants to pay them decently too. At its Louisiana shipyard, Saronic is undertaking a $300 million expansion that’s projected to create 1,500 direct jobs at an average salary of more than $87,000, 46% above the average in St. Mary Parish, where it’s based. Evan Boudreaux, the economic development director for the parish, has called it a project with “a generational impact.”The company’s next big bet is an AI-enabled shipyard that may land on Texas’s Gulf Coast, not far from SpaceX’s Starbase. According to the company, the proposed $3.2 billion project, called Port Alpha, could create up to 10,000 jobs and tap into the area’s rich talent pool. Once completed, Port Alpha would be twice the size of the country’s largest shipyard and capable of manufacturing everything from fast-attack boats to cargo ships. Last month, county commissioners approved a 95% tax break for the company, pending investment and hiring milestones. “We’re reshaping the entire shipbuilding industry now,” Mavrookas says.Saronic’s if-you-build-it-they-will-come bet will need more than just deep-pocketed investors. Reviving the maritime industrial base will require alliances between Washington and the private sector. Mavrookas points to 1980s South Korea, when the government marshaled a wave of financing and policy to shore up its shipbuilding industry. “It takes industry, government, and private capital to come together in a way that hasn’t happened (in the U.S.) in the past,” he says. “But we’re seeing the right moves from the administration.” As part of a renewed emphasis on shipbuilding, the second Trump administration has even sought Korea’s expertise. Through the Make America Shipbuilding Great Again (MASGA) initiative, three Korean companies—Hanwha, HD Hyundai, and Samsung Heavy Industries—have committed tens of billions of dollars to building ships on American soil. And across the Pentagon, efforts to reform rusty acquisition processes and grow a U.S. drone industry have given defense startups more confidence they can compete with giant prime contractors like Lockheed Martin. (It also helps that Saronic has a strong bench of seasoned ex-officials as advisors, including former Navy secretary John Lehman, who is also the uncle of one of the company’s cofounders, Rob Lehman.) Even so, some naval experts remain skeptical. The Navy is still working out how best to deploy drones, how to transport and service them, and how to ensure they can reliably talk to one another at a distance and when communications are jammed—a challenge Mavrookas admits is “very real.” Until those issues are addressed, chief of naval operations Admiral Daryl Caudle recently said, a robot boat is just “a PowerPoint-deep gadget and it doesn’t bring any real Navy combat power.”Mavrookas says Saronic is solving those technical challenges. To ensure its systems can work in the rigors of the sea, the company conducts thousands of miles of ocean testing of its vessels each month. It also makes a point of sitting designers next to builders and engineers alongside sailors in its offices, creating a tight feedback loop that helps the company move a software update from simulation to hardware in 30 minutes and into ocean trials within 24 hours. These tests, meanwhile, generate terabytes of training data—important for ensuring computer vision models can distinguish, say, a warship from a fishing vessel, or simply avoid collisions in busy waterways. Last month, a Royal Navy USV collided with a racing yacht during a training exercise within Portsmouth Harbor. (Photo: Saronic)Though Mavrookas won’t comment on specific tests—including a Navy trial last summer in which another company’s USV flipped the manned boat that was towing it—he claims Saronic has never had a safety incident caused by a malfunction in one of its platforms. But he also says that failures are an essential part of the process when you’re moving fast. “If you’re not testing systems in realistic environments and learning quickly from what works, what doesn’t,” he says, “you simply aren’t testing aggressively enough.”Mavrookas also avoids discussing the “lethality” of his products, unlike the founders of defense-tech giants Palantir and Anduril. He’s building fleets of drone boats not to go to war, he says, but to deter it. “When potential adversaries know the U.S. can rapidly generate and sustain maritime presence, replace losses, and protect critical waterways, the threshold for aggression rises,” he says. Still, it’s not hard to imagine how swarms of roboboats, prowling for suspicious activity, could also amplify tensions or provoke conflict. Mavrookas counters that as the world’s waters become increasingly contested, autonomous vessels offer operators more awareness, enabling better decision making. “And then, if you are in a conflict,” he says, “how do you keep sailors out of harm’s way and deliver real capabilities?”That remoteness carries its own risks. By lowering the economic and moral costs of war, and offloading human judgment, critics say AI-enabled targeting is already leading to more civilian casualties.As someone who has participated in life-or-death missions, Mavrookas sees things differently. “Would you rather have a naval mine floating in the ocean, completely indiscriminate? Or would you rather have a smart platform with computer vision that can make intelligent decisions based on the directions of a human?”But what will those directions look like? The Pentagon’s official policy on autonomous weapons doesn’t require human control; instead drone operators must merely “exercise appropriate levels of human judgment over the use of force.”Deciding what’s appropriate for a drone to do is a much more complex challenge than creating the software that empowers it or even the shipyard that builds it. And this one is far above Mavrookas’s pay grade. “It is very much a policy decision,” he says. “It’s not a technology decision.”
Diterbitkan : 2026-07-01 11:00:00
sumber : www.fastcompany.com
