Production Capacity Is the Binding Constraint on Defense Growth

The “next trillion-dollar category” frame does not get a market-size forecast here. The substance points elsewhere: defense-relevant companies can only become much larger if the United States can build systems fast enough, cheaply enough, and at enough scale to meet a new era of competition.

The two central claims come from different positions in that system. Dino Mavrookas, building Saronic and describing plans for Port Alpha, speaks from the company and manufacturing side: autonomy changes what can be built, how it can be built, and who can build it. Michael Duffey, speaking from a Pentagon acquisition and industrial-base perspective, treats production capacity as a binding constraint and says the Department of Defense has to change how it works with industry.

Their arguments converge on a narrow point: defense cannot scale if it remains dependent on bespoke platforms, fragile suppliers, scarce labor, and production rates calibrated only to annual defense budgets. The opportunity depends on rebuilding the industrial base around speed, producibility, private capital, and commercial demand.

Dino Mavrookas frames autonomy as a way to avoid sending people into combat when a robot can be sent instead. Warfare, he says, will still require human involvement “at some point,” but the premise for autonomous systems is direct: protect human life where possible.

That human-protection argument leads to an industrial one. For Mavrookas, autonomy matters because it changes the design space. If a company is merely promising to build the same ships the Navy has built for decades, or the same kinds of ships being built in China, only “better, faster, and cheaper,” the secretary should “politely ask me to leave.” The work is not incremental shipbuilding improvement. It is to “bend the economic cost curve.”

Mavrookas reduces that curve to two levers: material costs and labor costs. The United States is not going to buy steel more cheaply than China, and it is not going to acquire shipyard labor at a lower hourly rate. The alternative is to design ships that use less steel and require far fewer labor hours. That means building around software, autonomy, and digitization from the beginning rather than treating them as add-ons to legacy platforms.

He offers Saronic’s Marauder as an order-of-magnitude example, while explicitly cautioning that the comparison is “apples and oranges.” The first Marauder, he says, uses about 50,000 labor hours, compared with 7 million to 9 million labor hours for a destroyer. The comparison is not offered as platform equivalence. It is meant to show how autonomy can allow a platform to be redesigned so it is much simpler to build.

The military effect follows from the production effect: if autonomous ships can be made cheaper and faster, the Navy can field them at a scale that keeps more people out of combat scenarios. Autonomy, in this account, is not just a capability inside the vessel. It is part of the reason the vessel can be smaller, simpler, and more producible in the first place.

The maritime industrial base does not only lack capacity; it lacks time. Dino Mavrookas says experienced shipyard labor cannot be conjured on demand: “You can't create people with 15 years of experience overnight.” His answer is not to deny the workforce gap, but to design ships and processes so that workers do not need 15 years of welding experience to be productive.

He describes this as another first-principles design problem. If a worker coming from Ford, General Motors, Boeing, or SpaceX cannot become highly effective at building ships quickly, “that’s our fault.” In that case, the ship design is not simple enough, or the company lacks adequate processes, work instructions, training materials, or some other enabling system.

The line he borrows from Saronic’s head of manufacturing captures the intended shift: “less like an encyclopedia, more like IKEA.” The ambition is to make shipbuilding legible enough that industrial labor from adjacent sectors can be retrained into maritime production. Mavrookas ties that to culture as well as process: rebuilding what it means to work at a shipyard, making it “cool,” giving workers a mission, and giving them job security.

For him, workforce development is not separable from product architecture. A difficult-to-build design preserves dependence on scarce experience. A simpler design expands the labor pool.

Michael Duffey identifies production as the central bottleneck in the defense system, particularly for foreign military sales. Production rates, as he describes the problem, are generally calibrated to the size and cadence of the U.S. Department of Defense budget. That model is no longer sufficient.

Duffey says deals made with traditional industry “in the last two months” have tried to change the incentive structure by encouraging companies to spend their own private capital to expand production. He contrasts that with the traditional model of government support as a “handout.” The desired shift is for industry to take ownership of efficiency and production capacity rather than waiting for government-funded expansion.

That matters for both established contractors and startups. Duffey says the approach can “level the playing field” for new startups that are already operating with modern manufacturing and private-capital assumptions, while also bringing traditional defense contractors with needed capabilities into the same construct. His claim is not that legacy firms are dispensable, but that production capacity and efficiency need to be treated as industry-wide responsibilities.

This is where his position aligns closely with Mavrookas’s. Mavrookas emphasizes platform design and manufacturing simplification; Duffey emphasizes acquisition structures and incentives. Both treat production as more than a back-office concern. It is a condition for delivering military capability on schedule.

For Dino Mavrookas, Saronic’s Port Alpha is meant to support more than defense contracts. He describes it as a “generational project” and says the company is looking at building “one of, if not the largest shipyard in the world,” focused on autonomous platforms and the future of the maritime industry.

The commercial market is central to that plan. Mavrookas names larger vessels, cargo containers, bulk carriers, and oil tankers, linking them to broader goals such as energy dominance. The commercial logic is not a side business in his account; it is what allows the company to provide wartime production capacity during peacetime. Mavrookas phrases that customer as the “Department of War”; elsewhere in the exchange, the institutional context is the Department of Defense.

Without a commercial base, he warns, a defense-focused company ends up waiting for “the next contract vehicle and the next paycheck” to sustain itself. That may work for five, ten, twenty, or thirty years, but eventually the system could return to the same fragile structure it is trying to escape. A viable commercial business gives the industrial base a reason to stay scaled when the government is not actively buying at wartime levels.

Michael Duffey reinforces the same point from the Pentagon side. “Commercial first,” he says, is a major element of the acquisition transformation strategy because it improves resilience. He describes the traditional defense industrial base as frustratingly fragile: a sole supplier, not especially profitable, built around a bespoke defense design, becomes a vulnerability created by the government’s own procurement choices.

Erin Price-Wright summarizes the premise with a line her team often uses: “there’s no defense industrial base without an industrial base.” That is not a slogan about reshoring in general. It is a specific claim about dependence: defense capacity cannot be resilient if the underlying commercial and manufacturing base is too thin.

Duffey’s message to Pentagon program executive officers, program managers, and founders is that acquisition transformation cannot succeed through a single central reform. It depends on letting “a thousand flowers bloom” while removing obstacles that prevent companies and program leaders from delivering capability to the warfighter on schedule.

That places responsibility on the bureaucracy as well as on industry. Michael Duffey thanks leaders who are “fighting through the bureaucracy,” but he also says the department is trying to change culture inside the Pentagon through “maximum communication” across the acquisition hierarchy and directly with industry. The operational request is specific: people should identify what is not working and bring those barriers to his team or others, rather than waiting for reform to arrive from the top.

The tone is incremental rather than triumphal. Duffey says the work will take time and cooperation. But he presents the direction clearly: faster delivery requires more direct communication about obstacles, more willingness to remove them, and a broader set of companies able to help deliver capability.

Dino Mavrookas closes with a builder’s version of the same appeal. The country faces a “real generational opportunity” to build what it needs for the next hundred years, but no single company or person will solve the problem. The need, in his formulation, is for more founders, more builders, and more people in government pushing for change.