Defense tech’s next test is not software but factories, shipyards and scale

A new phase of the defense technology cycle is beginning to emerge in the United States. Companies that built their reputations around autonomous systems, artificial intelligence, drones and faster weapons development are increasingly committing large amounts of private capital to factories, shipyards, aerospace production and heavy industrial capacity. The change suggests that challenging the traditional defense establishment may ultimately require many of the same physical capabilities that defined it: manufacturing space, skilled workers, qualified suppliers, production engineering and the ability to deliver hardware repeatedly at scale.

Anduril Industries is preparing a $3.7 billion shipbuilding investment in Maryland while simultaneously producing autonomous combat aircraft at its newer Ohio manufacturing complex. Saronic Technologies has broken ground on a shipyard representing more than $3 billion of planned private investment in Texas. Hadrian has opened a major automated submarine-component facility in Alabama and secured another $1.37 billion of equity financing to expand its manufacturing network, while Shield AI has chosen Washington and Kansas for production, flight testing and sustainment of its planned X-BAT aircraft.

These developments do not yet prove that venture-backed defense companies are becoming full replacements for established prime contractors. Much of the largest capacity remains under construction, some programmes will not reach production for several years, and decades of evidence from U.S. naval procurement show that industrial complexity can defeat even experienced manufacturers. What is becoming visible, however, is a broader structural shift in which defense technology competition is moving from software laboratories and prototype demonstrations onto factory floors.

Why are defense technology companies suddenly investing billions in physical production?

The economics of defense technology change once a company moves beyond small production runs. Software can be replicated quickly, but ships, aircraft, missiles, propulsion systems and precision components require facilities, equipment, trained technicians and supplier networks. Winning a prototype competition therefore solves only the first part of the commercial problem.

Anduril Industries provides perhaps the clearest example of that transition. On October 6, the company announced Arsenal-2 at Sparrows Point in Baltimore County, Maryland, where it plans to invest $3.7 billion in a roughly 2-million-square-foot manufacturing operation focused on critical components for Virginia-class submarines. The U.S. Navy has separately awarded Anduril a contract worth up to $2.9 billion, with the company saying payments will be linked to demonstrated production outcomes.

The combined commitments put as much as $6.6 billion behind a company founded in 2017 and better known for autonomous weapons, surveillance technology and software than nuclear-submarine manufacturing. Permitting and site preparation are beginning first, with initial operations currently expected around 2030 and production intended to scale thereafter.

The distinction between announced investment and operating capacity is important. Arsenal-2 will not relieve submarine production constraints immediately, and manufacturing critical components for nuclear-powered submarines introduces regulatory, quality and supply-chain requirements very different from building autonomous drones. Yet the size of the commitment shows how far the company intends to move beyond the conventional image of a defense technology startup.

Is Anduril proving that newer defense companies can move from prototypes into production?

Anduril’s Ohio operation provides more tangible evidence because manufacturing has already started. Arsenal-1 was announced in early 2025 as a planned hyperscale manufacturing complex near Columbus, with company investment approaching $1 billion and a long-term footprint designed to reach around 5 million square feet.

The facility produced its first Ohio-built Fury autonomous combat aircraft in July 2026, approximately 18 months after the site was announced. Production work had started months earlier, providing an early indication that the company could move at least one programme from factory construction into hardware output relatively quickly.

That does not establish sustained production performance. One completed aircraft is different from maintaining quality, delivery schedules and cost discipline across dozens or hundreds of systems each year. The more important milestone is that the manufacturing thesis has moved beyond architectural renderings and planned factory capacity into an operating production line.

Anduril’s broader industrial footprint also stretches beyond Ohio. The company has developed manufacturing capabilities across solid rocket motors, autonomous underwater vehicles, launched systems and other defense products, making Arsenal-2 less isolated than it might initially appear.

The strategic direction is therefore becoming clearer. Rather than remaining dependent on incumbent manufacturers to turn its designs into physical systems, Anduril increasingly appears to be building the industrial infrastructure needed to control a greater share of design, production and delivery itself.

Could Saronic transform autonomous-boat expertise into large-scale American shipbuilding?

Saronic Technologies is following a similar path in the maritime sector. The company broke ground on Port Alpha at the Port of Brownsville in Texas on September 30, advancing a project involving more than $3 billion of planned private investment.

The site initially covers 835 acres and includes the option for significant future expansion. Saronic plans an initial phase with approximately 1.2 million square feet of enclosed production space, multiple dedicated production buildings and deepwater access. Initial operations are targeted for 2028.

The company’s ambitions are notable because Port Alpha is not being designed only for small autonomous vessels. Saronic says the shipyard will be capable of producing autonomous, autonomy-capable and crewed vessels, with initial construction intended to include Landing Craft Utility vessels for the U.S. Navy. The company says initial annual capacity could reach 150,000 gross tons of shipbuilding output, although that remains a forward-looking target until the facility is operating.

This represents a very different industrial proposition from developing unmanned surface vessels. Building larger naval and commercial vessels requires steel processing, large-scale assembly, waterfront infrastructure, workforce development and extensive supplier coordination. Saronic is effectively betting that principles associated with modern software and advanced manufacturing can be applied to a sector where capacity expansion traditionally takes years.

Whether that works at the proposed scale remains uncertain. The significance for the wider industry lies in the willingness of private capital to attempt it at all. A defense technology company is no longer merely seeking a place inside the existing U.S. shipbuilding supply chain. It is trying to create substantial new shipbuilding capacity.

Why is automated manufacturing becoming as important as new weapons platforms?

Hadrian represents another part of the same structural change because its business is manufacturing capacity itself. Rather than competing primarily by designing a new fighter, ship or autonomous platform, the company is building highly automated factories intended to produce parts, assemblies and eventually more complete systems for aerospace and defense customers.

In March 2026, Hadrian opened Factory 4 in Cherokee, Alabama, dedicated to supporting Columbia-class and Virginia-class submarine production. The public-private programme combines more than $1.5 billion of private capital with $900 million in U.S. Navy funding, taking the total planned investment above $2.4 billion.

The facility occupies part of a roughly 2.2-million-square-foot industrial site and is intended to manufacture submarine components, assemblies and finished products. Hadrian says the project is designed around automation intended to reduce the dependence on lengthy traditional training cycles for some manufacturing tasks, an important claim in an industrial base where skilled labour shortages have repeatedly constrained production.

The company then secured $1.37 billion in new equity financing in August at a stated valuation of $7.87 billion. Hadrian said the capital would support additional factories, research and development and new production capabilities across areas including munitions, shipbuilding and autonomous systems. At that point, the company reported a manufacturing footprint of just under 3 million square feet across four sites.

The financing provides another indication that private capital is treating defense manufacturing capacity as a growth market rather than solely a government-funded industrial activity. Hadrian’s challenge will be demonstrating that factory automation can produce the repeatability, quality and certification standards required for mission-critical defense hardware while meaningfully increasing throughput.

Is autonomous warfare demand creating enough procurement depth to support these factories?

Manufacturing expansion would be much less significant without a parallel change in government demand. The fiscal 2026 U.S. defense budget request identified $13.4 billion specifically for autonomy and autonomous systems, including unmanned aerial vehicles, maritime systems and underlying autonomy capabilities.

That budget treatment matters because autonomy is increasingly being recognised as a distinct procurement category rather than an experimental technology attached to conventional platforms. The Department of the Navy portion alone included billions of dollars across unmanned systems, while the broader request directed substantial funding toward aerial autonomous capabilities.

Organisational structures are shifting as well. U.S. Southern Command announced the establishment of an Autonomous Warfare Command in April 2026, intended to integrate autonomous, semi-autonomous and unmanned platforms into operational missions across multiple domains. The initiative remains relatively new, but it adds to evidence that autonomy is moving from demonstration programmes toward regular military force structures.

This creates a reinforcing mechanism. Larger procurement programmes justify factory investment, while greater factory capacity can make higher-volume procurement more plausible. If autonomous systems are expected to be fielded in quantities materially larger than traditional exquisite platforms, production architecture becomes part of the military capability itself.

The same principle applies to conventional platforms. The Navy does not merely need additional submarine designs. It needs enough industrial throughput to build the submarines already planned, while simultaneously supporting the Columbia-class ballistic missile submarine programme and obligations associated with the broader submarine industrial base.

Could manufacturing capacity become the real competitive advantage in defense technology?

The first generation of the recent defense technology boom focused heavily on speed of product development. New entrants argued that commercial software practices, faster iteration and private capital could shorten acquisition cycles that traditionally stretched across many years.

The next competitive test may be considerably harder. Manufacturing at scale requires a different operating discipline. Suppliers have to qualify components, workers must maintain production standards, factory utilisation has to increase without sacrificing quality, and delivery schedules must survive equipment failures, material shortages and design changes.

That creates an important strategic divide between companies capable of building a compelling prototype and companies capable of becoming durable industrial suppliers. The larger the procurement programme, the more the second capability matters.

It could also change how the defense industry is financed. Traditional prime contractors generally developed enormous manufacturing footprints over decades of government contracting. New entrants are attempting to compress parts of that process by deploying venture capital, growth equity and private investment into facilities before every production requirement is certain.

Anduril’s $3.7 billion private commitment to Arsenal-2, Saronic’s more than $3 billion Port Alpha programme and Hadrian’s substantial private contribution to its Alabama factory illustrate that shift. Private capital is increasingly financing industrial capacity that may ultimately depend on government procurement for a significant share of its utilisation.

If successful, this could expand the number of companies capable of competing for large defense programmes and reduce some capacity bottlenecks. If demand arrives more slowly than anticipated, however, capital-intensive facilities could become expensive assets searching for sufficient production volume.

Why does the U.S. Navy’s shipbuilding record make the startup manufacturing thesis harder to prove?

The strongest contrary evidence comes from the industry the new entrants are trying to disrupt. U.S. government assessments show that naval shipbuilding has experienced persistent schedule delays, cost growth and workforce constraints despite years of investment and the involvement of contractors with extensive shipbuilding experience.

The U.S. Government Accountability Office reported in 2026 that Virginia-class submarine construction was operating at roughly one submarine per year as of June 2025, only half the Navy’s goal of two annually. Although two Virginia-class submarines were delivered during 2025, both arrived more than three years late.

The Columbia-class programme has also faced schedule pressure, while other naval programmes have experienced significant delays. Those difficulties demonstrate why factory announcements alone cannot establish a manufacturing breakthrough. Shipbuilding combines design maturity, materials, workforce experience, supplier performance, quality assurance and programme management in ways that cannot necessarily be solved by software or capital.

Workforce availability remains another constraint. New facilities compete for welders, machinists, manufacturing engineers and technicians at the same time that incumbent shipyards and aerospace manufacturers are trying to expand. Automation may reduce some labour intensity, but it does not eliminate the need for deep technical expertise.

The risk therefore runs both ways. The shortcomings of the existing industrial base create an opening for new entrants, but those same shortcomings demonstrate how difficult the underlying problem is.

Is Shield AI another sign that software-led defense firms are becoming aerospace manufacturers?

Shield AI provides a useful test at an earlier stage of the production cycle. The company developed much of its identity around autonomy software and unmanned aircraft, but in September it selected Des Moines, Washington, as the production home for its planned X-BAT aircraft and Newton, Kansas, as the programme’s flight-test, acceptance and sustainment centre.

The Washington site is being built out, hiring has begun and Shield AI says X-BAT production is planned to start in 2029, with full-rate production targeted for the early 2030s. Every aircraft is expected to undergo production acceptance testing in Kansas before customer delivery if the programme develops according to plan.

Those dates underline why the broader structural Signal should not be overstated. Shield AI has chosen an industrial architecture, but X-BAT is not currently being produced at scale. Saronic’s Port Alpha is under construction rather than operational, and Anduril’s Arsenal-2 remains years from initial operations.

Nevertheless, the direction of travel is consistent across companies that developed independently and operate in different parts of the defense market. Software-led and autonomy-led businesses are acquiring the facilities and organisational capabilities traditionally associated with aerospace and industrial contractors.

What would confirm that defense tech startups are becoming a new class of industrial prime?

The strongest confirmation over the next 6 to 24 months would come from production rather than investment announcements. Arsenal-1 will need to demonstrate sustained aircraft output after its first Fury milestone, while Hadrian’s Alabama operation will need to show that automated manufacturing can translate into measurable improvements in component availability and submarine schedules.

Saronic’s construction progress at Port Alpha will be equally important. Meeting the 2028 initial-operations target, establishing its workforce and supplier network and beginning contracted vessel production would move the project from ambitious industrial plan to functioning shipyard. Material delays or reductions in planned capacity would weaken the thesis.

Government procurement will provide another test. Additional production contracts awarded to newer defense companies, especially contracts that progress from prototype work into recurring quantities, would demonstrate that the industrial expansion has sufficient demand behind it. Autonomy spending will matter most when appropriated money converts into delivered systems rather than programme announcements.

Capital deployment also needs watching. Further multibillion-dollar private investments in munitions plants, aerospace factories, shipbuilding facilities and defense supply chains would suggest that institutional investors increasingly view production capacity as a scalable defense technology opportunity.

The ultimate measure will be whether these companies can deliver hardware consistently enough to change procurement behaviour. Traditional prime contractors retain enormous advantages in certification, engineering knowledge, customer relationships, supply chains and experience managing complex military programmes. Those advantages will not disappear because newer companies construct impressive factories.

What is changing is the competitive boundary. The defense technology challengers that once differentiated themselves primarily through software, autonomy and development speed are increasingly investing in the industrial capabilities required to manufacture at scale. If that transition succeeds, the next generation of major defense contractors may not emerge simply by writing better software or designing better drones. It may emerge from companies that learn how to combine those capabilities with the much older disciplines of factories, shipyards, supply chains and production.