For most of the 20th century, American power rested on a simple industrial premise: when the stakes were highest, the United States could build more, faster, and better than anyone else.

Today, that premise is no longer safe to assume.

America still leads in invention. We have the deepest capital markets, the strongest software ecosystem, the leading research universities, and many of the world’s most advanced defense platforms. But over the last four decades, we allowed the industrial layer beneath that innovation to hollow out. We did not simply offshore low-cost assembly. We offshored tooling, process knowledge, supplier depth, workforce training, and the hard-earned manufacturing expertise required to turn a design into a qualified physical product.

America can still design advanced missiles, autonomous drones, satellites, submarines, electric motors, and AI infrastructure better than our near-peer adversaries. The harder question is whether we can source the materials, machine the parts, scale the factories, and replenish the systems fast enough in a crisis – particularly when the window of relevance is measured in weeks or days.

The warning signs are impossible to ignore. Since the 1990s, the U.S. aerospace and defense prime contractor base has consolidated from 50+ companies to just five, leaving the Department of War dependent on a narrower set of suppliers for critical capabilities. Beneath those primes sits an even more fragile base of small machine shops, specialized component suppliers, aging equipment, and scarce skilled labor. The result is a defense ecosystem with world-class engineering at the top and brittle production capacity underneath.

This leaves the United States in an uncomfortable position: in too many categories central to national and economic security, we now rely on supply chains that run through the very country we are preparing to deter.

Rare earth magnets are one of the clearest examples. They sit inside fighter jets, guided missiles, drones, satellites, electric motors, robotics, and AI data infrastructure. Yet in 2024, China produced roughly 300,000 metric tons of NdFeB rare earth magnets, while the United States produced fewer than 1,000. China controls roughly 69% of rare earth mining, 92% of refining, and 98% of magnet manufacturing. This is not a normal supply-chain imbalance. It is a strategic dependency embedded inside the hardware layer of the modern economy.

And that dependency is already being exposed. Since April 2025, China has tightened export licensing for rare earth and magnet-related materials, introduced end-user tracking requirements, and warned trading partners against facilitating indirect exports. China’s restrictions have already upended supply chains for U.S. automakers, aerospace manufacturers, semiconductor companies, and military contractors.

A leading great power should not depend on a near-peer adversary for components that determine whether aircraft, missiles, satellites, drones, autonomous systems, and AI infrastructure can be built at all.

The same pattern runs through the broader industrial base. Decades of offshoring and underinvestment have left U.S. manufacturers dependent on fragmented supplier networks, long logistics chains, scarce skilled labor, and small domestic shops that often lack the capital or digital infrastructure to scale. Much of the domestic contract manufacturing base still operates with aging equipment and limited software-enabled production infrastructure, while roughly 400,000 manufacturing roles remain unfilled.

Why This Moment is Different

America’s industrial base has weakened while the demands placed on it are accelerating.

Missile defense, autonomous systems, shipbuilding, space architectures, directed energy, power infrastructure, and AI compute all require more trusted production capacity, not less. At the same time, modern conflict is moving faster than traditional industrial systems were designed to support. Ukraine has made the lesson impossible to miss: battlefield advantage increasingly belongs to the side that can replace, modify, and redeploy physical systems continuously. Manufacturing speed is becoming part of the kill chain.

A Taiwan scenario would expose the same issue on a larger scale. Multiple war games suggest the United States could run out of key long-range precision-guided munitions in less than one week in a Taiwan Strait conflict. America can build the most advanced systems in the world but the question is whether it can build enough of them, replenish them, and adapt them under pressure. We have a true production gap problem in our defense industrial base.

This is why American reindustrialization cannot mean recreating the old industrial base exactly as it was. The old model was too centralized, too labor-constrained, too slow to retool, and too dependent on fragile global supply chains. The answer must be a new industrial stack: secure materials, software-defined production, automated factories, and manufacturing capacity that can scale from prototype to production before the opportunity – or the threat – has moved on.

Industrial capacity is no longer a background condition of national power. It is one of the decisive constraints on it.

What Reindustrialization Has to Become

The answer is not simply more factories. It is a different kind of industrial base.

The new industrial base must be built around four layers.

First, secure inputs. If critical materials, magnets, electronics, castings, forgings, or precision components depend on adversarial supply chains, the rest of the industrial base is only conditionally useful.

Second, software-defined production. The next generation of factories must be able to ingest designs, automate engineering workflows, generate machine instructions, inspect output, and learn from every production run. The factory can no longer be a static asset. It must become a learning system.

Third, automated factory platforms. The country cannot solve a production crisis with labor models that depend on too few skilled workers and low-utilization equipment. We need manufacturing systems that improve throughput, quality, and utilization while reducing dependence on labor bottlenecks.

Fourth, distributed capacity. Future conflicts will require not only centralized production, but also the ability to repair, modify, and regenerate systems closer to the point of need. Industrial capacity that cannot reach the mission in time may not matter when the mission arrives.

The common thread is speed. The winners in this new industrial era will not simply make parts domestically. They will compress the time between design and qualified output.

Why New Companies Can Win

The incumbent defense industrial base remains essential. The primes will continue to integrate major systems, manage complex programs, and serve as critical channels to the Department of War. But the primes were not built to reinvent the production layer from first principles.

Their operating model was shaped by long-cycle programs, compliance-heavy procurement, bespoke production runs, and cost structures not designed for rapid iteration or unit-cost compression. Many legacy suppliers beneath them are even more constrained: too small to access meaningful growth capital, too labor-dependent to scale quickly, and too tied to aging equipment to become software-native through incremental upgrades.

That creates room for a new category of industrial company. These companies are infrastructure platforms. They own the workflow, the data, the software, the factory architecture, and in some cases the critical materials layer itself. They are designed to compound: every customer, part, facility, production run, and inspection cycle improves the system.

This is why reindustrialization is investable. There is an opportunity to bring production back to the United States and to build companies that can produce with better unit economics, faster learning curves, and greater resilience than the legacy base they are replacing.

Who We’re Backing

Washington Harbour Partners is actively investing across this new industrial stack.

Vulcan Elements is rebuilding the materials layer by creating a China-free rare earth magnet supply chain. Its platform spans the full production lifecycle, from oxide inputs through finished magnet fabrication. With Washington Harbour Partners as the lead investor in the $1.4B private-public partnership deal from the Department of War and the Department of Commerce, the Company is now expanding toward a 10,000 metric ton facility for American-made and processed rare earth magnets.

Hadrian is rebuilding precision metal manufacturing as a software-defined factory platform. Its Opus system automates design ingestion, manufacturability analysis, machine programming, factory orchestration, and closed-loop inspection. With factories operating at 70%+ machine utilization versus a 15% industry average, Hadrian is changing both the speed and unit economics of defense industrial production.

Mariana Minerals is rebuilding critical-minerals production around its proprietary, AI-powered software platform, MarianaOS, creating a software-defined and autonomous mining and refining company. Mariana acquires, develops, and operates its own mines and refineries, deploying MineOS to autonomously orchestrate mining operations, PlantOS to optimize refinery processes through closed-loop reinforcement learning, and CapitalProjectOS to accelerate engineering and construction. The Company’s ambition is to make new Western minerals capacity faster and cheaper to build, ramp, and operate across copper, lithium, and ultimately a much broader set of strategic materials.

VulcanForms is building factory-scale advanced metal manufacturing for complex components that are difficult or uneconomic to produce through traditional methods. Leveraging its proprietary laser powder bed fusion systems that use 150 lasers per printer for additive manufacturing vs. 1-4 in conventional systems, with production lines designed for 24/7 operation, we are seeing a step-function change in America’s additive manufacturing base.

Firestorm is bringing a new model of manufacturing by extending industrial capacity to the tactical edge. The Company’s xCell platform enables forward-deployed manufacturing of 3^rd party systems, government-owned IP, legacy parts, and mission-critical components that otherwise face long depot or OEM lead times. Future conflicts will require not just centralized production capacity, but the ability to deploy manufacturing systems dynamically in contested environments. Firestorm is built for that future.

Let’s Reindustrialize America

These are all different layers of the same thesis: the next era of American power will depend on who has the best technology, and on who can manufacture what the mission requires at the speed the moment demands.

Reindustrialization is not about returning to the past. It is about building the production architecture required for the next century of American power.

At Washington Harbour Partners, we believe this new industrial stack will produce some of the most important companies of the next decade – and we are focused on backing the builders turning industrial capacity from America’s constraint into America’s advantage.