<![CDATA[American Forge Institute: Sunday Spotlight]]>https://lukeringlein.substack.com/s/spotlight-sundayhttps://substackcdn.com/image/fetch/$s_!n_jJ!,w_256,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F52417c87-981e-4296-86a3-18a9e4323420_500x500.pngAmerican Forge Institute: Sunday Spotlighthttps://lukeringlein.substack.com/s/spotlight-sundaySubstackSat, 30 May 2026 09:06:19 GMT<![CDATA[Spotlight: America’s Next Mine Might Be a Million Miles Away (AstroForge)]]>https://lukeringlein.substack.com/p/spotlight-americas-next-mine-mighthttps://lukeringlein.substack.com/p/spotlight-americas-next-mine-mightSun, 17 Aug 2025 15:49:29 GMT

The AFI posts briefs, spotlights, and news every week. All of our content is completely free, forever. If you care about American industry and national security, consider joining the movement.

Join the movement

Company Fundamentals

AstroForge opened shop in January 2022. It was started by Matt Gialich, who cut his teeth as an engineer at Virgin Orbit, and Jose Acain, a former deputy chief engineer at SpaceX. They set up in Huntington Beach, not Silicon Valley. The mission was blunt from the start: mine platinum-group metals from asteroids and cut out the toxic mess of digging them out of Earth. No hype about space fuel depots or moon colonies, just metals with an actual market. That focus hasn’t shifted, only the tempo. The company decided to move faster, even if that meant higher odds of failure.

Money came quickly. A seed round in 2022 pulled in $13 million, led by Initialized Capital. By 2024 they had closed a $40 million Series A, with investors like Alexis Ohanian’s Seven Seven Six, Y Combinator, and Jed McCaleb backing the gamble. In total, more than $50 million is on the line. Gialich doesn’t soften it: “We’re going to mine asteroids or go bankrupt.” The team grew from about twenty people in 2023 to around 35 by mid-2024, adding spacecraft engineers after early mistakes exposed gaps. One milestone already checked off: AstroForge secured the first FCC license ever to run a commercial spacecraft beyond Earth orbit.

The company works out of its Huntington Beach lab, where most of the engineering happens. Early hardware came from partners like OrbAstro, but after Brokkr-1 failed, they started pulling more work in-house. Specialized testing still leans on outsiders, like Safran firing their thrusters in France. And AstroForge is making sure it’s wired into the industry’s power circuits. In 2025, they joined the Commercial Spaceflight Federation, putting themselves in the room where space policy and lobbying collide. It’s a small team building a big supply chain from scratch, with the clock running.

Technology and Missions

Brokkr-1

AstroForge’s first try went up in April 2023 on a Falcon 9 rideshare. The 6U cubesat carried a pinch of asteroid-like rock meant to prove their refinery tech could pull metals out in microgravity. The plan was to vaporize the ore and condense platinum. It never got that far. Contact with the satellite dropped almost immediately, likely from a bad comms link, and the experiment never fired.

The mission cost them money and pride, but it showed they were serious about flying hardware instead of polishing slide decks. As Matt Gialich told reporters: “We know we’re gonna have some failures…and we welcome them. That’s how we learn.” Lesson one was obvious: build tougher radios and test harder on the ground.

Odin (Brokkr-2)

Next came Odin, launched in February 2025. This probe was supposed to swing past a 100-meter asteroid called 2022 OB5, suspected to be a nickel-iron body. The craft would take images and spectral readings to prove the rock was worth mining.

Two days in, everything went sideways. A ground station amplifier blew, crippling communications. They managed to ping the probe a few times, enough to know its solar panels had deployed, but that was it. By early March, contact was gone. The spacecraft itself had been rushed—rebuilt in 212 days after its first frame failed vibration tests—and Gialich had only given it a 30% chance of success from the start. Odin cost about $6.5 million all-in, peanuts compared to NASA missions, but still painful for a startup.

The loss forced upgrades: more robust wiring, redundant comms, and a push to hire seasoned spacecraft engineers, not just rocket guys. As Gialich admitted on the company blog, “We all have a lot of rocket experience...but this isn’t a f**ing rocket.”*

Vestri

Now comes Vestri, the big gamble. Scheduled to fly on Intuitive Machines’ IM-3 lander mission in late 2025 or early 2026, it will be the first private spacecraft to attempt an asteroid landing. At roughly 200 kilograms, it’s double Odin’s size and built to dock with 2022 OB5. The trick is magnets. If the asteroid is truly nickel-iron, Vestri will clamp down without needing harpoons or drills.

AstroForge partnered with Safran for advanced Hall-effect thrusters, giving Vestri the endurance to dance around the rock and attempt a controlled landing. The spacecraft will carry sensors to measure the alloy content and look for platinum, palladium, rhodium, and iridium. If it works, AstroForge will not only land on an asteroid but prospect it for the first time in commercial history.

One more mission is penciled in after Vestri: bring metal back to Earth. But that only happens if this one sticks the landing.

Asteroid Targets and Rationale

AstroForge is betting on M-type asteroids (the leftover cores of planets), packed with nickel, iron, and sometimes the platinum-group metals that industry craves. NASA missions have mostly chased C-types, rich in water and carbon. Water’s useful if you’re fueling space stations, but there’s no market for it today. Platinum sells. That’s why AstroForge chose metals over ice.

Platinum-group metals (platinum, palladium, rhodium, iridium) are rare, hard to replace, and painfully expensive. They’re in catalytic converters, fuel cells, electronics, and high-temp alloys. As CEO Matt Gialich has put it, mining them on Earth is dirty business, and the U.S. doesn’t have much of its own supply. Pulling them from space could both feed industry and cut down on the tailings ponds and acid leaching we live with now.

Why nickel-iron rocks specifically? Because the math works. Even a 100-meter metallic asteroid could hold tens of tons of platinum, worth billions on Earth. That’s not fantasy; NASA’s Psyche mission is headed for a much bigger metallic world, and speculative estimates put its metal value beyond comprehension. Those numbers are cartoonish, but they prove the point: a single asteroid can dwarf annual global output.

AstroForge’s first target is 2022 OB5, a 100-meter near-Earth asteroid flagged as possibly metal-rich. They’re cagey about the rest, but say they’ve identified five more candidates. The sweet spot is small, accessible bodies, 20 to 300 meters wide, because they’re easier to land on and leave. Only a few percent of near-Earth asteroids fit that profile, but that still leaves dozens to choose from.

If OB5 turns out to be a dud, AstroForge can pivot. Other likely prospects include 1986 DA and 2016 ED85, both thought to be loaded with metal. The company has already worked with universities and observatories to quietly gather spectra on candidates before announcing its target. That’s old-school prospecting, just with telescopes instead of picks.

The roadmap is simple: prove refining works in orbit (Brokkr-1), scout a real asteroid (Odin), land and analyze it (Vestri), then bring material home. Each mission builds pressure for the next. If Vestri clamps onto OB5 and confirms it’s rich in platinum, AstroForge won’t just have a target; it’ll have the first claim to off-world paydirt.

Industrial and Strategic Position

AstroForge isn’t the first outfit to talk asteroid mining. A decade ago, Planetary Resources and Deep Space Industries made noise about pulling water and minerals out of rocks. Neither reached an asteroid before running out of cash. They were ahead of the market, chasing rocket fuel no one wanted to buy yet, and stuck with tech too expensive for the time.

AstroForge is playing it differently. They’re going after metals with value today, not speculative markets. Platinum-group metals already move billions of dollars a year. And instead of betting on giant, bespoke spacecraft, they’ve kept costs down by riding Falcon 9 rideshares and leaning on smallsat tech. Brokkr-1 and Odin each flew for under $10 million—a sum that would barely cover paperwork at NASA.

Their strategy is speed. Launch, learn, fix, repeat. Gialich put it plainly: “We need to go fast, and we’re willing to take more risks.” Planetary Resources and DSI never managed that cadence.

There’s also focus. AstroForge has built its own in-space refining process, something the earlier players never demonstrated. It isn’t trying to be everything at once—no grand orbital infrastructure plans—just a lean pipeline: identify rock, dock, extract metal. Where it makes sense, they buy proven hardware. That’s why Safran is supplying Vestri’s thrusters, instead of the team wasting years reinventing propulsion.

Being small doesn’t mean being cut off. They went through Y Combinator early on, and in 2025 joined the Commercial Spaceflight Federation to keep a foot in policy circles. No direct NASA or ESA partnerships yet, but AstroForge benefits from the trail those agencies cut with missions like OSIRIS-REx.

The company’s pitch blends industry and national security. More platinum from space means less dependence on South Africa or Russia. Cleaner supply chains mean less destruction on Earth. That story sells with investors hunting for “hard tech” bets that matter. What AstroForge hasn’t done yet is prove it works. But compared to the last generation of asteroid miners, they’re already further along—flying real missions, learning the hard way, and keeping investors close.

Economics and Market

Platinum-group metals are the target. The world burns through millions of ounces of platinum every year, most of it dug out of South Africa. Palladium and rhodium follow the same story; concentrated supply, unstable politics, constant labor strikes. Prices jump with every disruption. When Russia pulled back palladium exports in 2022, automakers had to scramble for substitutes. AstroForge argues it can break that chokehold.

The metals also have growing uses beyond cars. Platinum shows up in fuel cells, hydrogen projects, and advanced electronics. Iridium is essential for making OLED screens and space-grade alloys. These aren’t optional luxuries; they’re the guts of the energy and defense industries. The U.S. classifies PGMs as critical minerals because a supply cut could ripple through the economy.

On paper, asteroid mining is worth trillions. A single metallic asteroid could hold more PGMs than Earth mines in a year. But AstroForge isn’t selling fantasy. The real economics start with mission cost versus cargo value. Odin flew for about $6.5 million. Vestri will likely run tens of millions. Even if they only return a few kilograms of platinum, that payload could clear costs at today’s prices.

Earth mining still wins on upfront expense, but it carries hidden costs: massive waste piles, acid runoff, carbon emissions. A refinery-in-orbit has no tailings pond to poison water supplies. If AstroForge can make returns at even twice the price of terrestrial mines, industries and governments may still pay for cleaner, more reliable supply.

Who would buy it? Automakers, electronics giants, and fuel cell suppliers all need steady PGMs. Defense contractors need iridium and rhodium for high-performance systems. The U.S. itself could be a customer. Washington already manages a National Defense Stockpile of critical minerals, and adding space-mined metals would fit its long push for supply security.

AstroForge doesn’t need to flood the market to matter. A few tons a year could stabilize prices, cut dependency on South Africa or Russia, and prove the concept. Scale comes later. First, they just need to show an asteroid isn’t just a rock, but a mine.

Challenges and Risks

Space doesn’t forgive mistakes. AstroForge already lost one satellite on launch and another to a fried ground station. Now they’re aiming to land Vestri on a metal rock whipping through space faster than a rifle round. NASA needed billions and decades to pull off a single asteroid sample return. AstroForge is trying it with a staff you could fit in a diner booth.

Regulation is another wall. The 2015 U.S. SPACE Act says American companies can own what they mine in space, but the Outer Space Treaty still bans nations from claiming celestial bodies. That legal gray zone hasn’t been tested. If AstroForge brings back platinum, expect lawsuits, diplomacy, and maybe new treaties before the bars are even poured.

Money is a constant risk. Investors have already put up more than $50 million. Vestri could burn through another big chunk. If it fails, raising fresh cash gets harder. Space is not like software—you can’t pivot to another app if the spacecraft explodes. The company’s own founder put it best: they’ll either mine asteroids or go bankrupt trying.

Strategic “Why It Matters”

For U.S. industry, AstroForge offers a way out of fragile supply chains. South Africa and Russia control the tap on most platinum-group metals. If they close it, automakers, chip fabs, and defense contractors feel the pain. A working asteroid mine would cut that dependence and lock a strategic resource inside American reach.

For the space economy, it’s more than metal. If AstroForge lands Vestri and proves the asteroid is rich, it will mark the first commercial step toward extracting resources beyond Earth orbit. That would turn “space economy” from PowerPoint slides into real supply lines. Failures are expected, but even partial success rewrites the playbook for how private firms operate in deep space.

For American dynamism, the story is grit. A small team in Huntington Beach has flown hardware, lost hardware, and is still doubling down. They’re not selling moon colonies or Mars tickets, just chasing a raw industrial input the country needs. If Vestri sticks, it’s proof that progress comes from builders willing to take the hit and get back up.

Thank you for reading AFI. For more free insight, stories, and news, join the newsletter!

Subscribe now

References

]]><![CDATA[Spotlight: Valar Atomics]]>https://lukeringlein.substack.com/p/spotlight-valar-atomicshttps://lukeringlein.substack.com/p/spotlight-valar-atomicsSun, 03 Aug 2025 20:14:28 GMT
Valar Atomics comes out of stealth with $19M and a pilot reactor site | TechCrunch

Forge Ahead is your go-to source for insight on America’s industrial future. Subscribe for FREE weekly news, company profiles, and strategic reccomendations.

Subscribe now

Company Origins and Mission

Isaiah Taylor dropped out of high school at 16 to write software for hedge funds. By his early twenties, he was pulling six figures and nursing a decade-long obsession with nuclear energy. His great-grandfather Ward Schaap worked on the Manhattan Project, so nuclear was literally in his DNA.

Taylor founded Valar Atomics in 2023 from El Segundo, California. The company sits in the heart of the emerging "Gundo Bros" deeptech cluster, Los Angeles's scrappy answer to Silicon Valley's polished venture scene. Taylor recruited Mark Mitchell as Chief Nuclear Officer, a 20-year industry veteran who previously led Ultra Safe Nuclear Corporation's reactor program.

Valar's mission sounds simple: replace nuclear's artisanal approach with mass production. Instead of building bespoke billion-dollar plants that take decades to deploy, they want to stamp out thousands of identical small reactors. The goal is abundant energy for AI data centers, industrial heat for manufacturing, and synthetic fuel production. Nuclear power that goes where it's needed, not where the grid allows.

Reactor Design and Technology

Valar builds High-Temperature Gas Reactors using TRISO fuel and helium coolant. The design can reach 900°C, three times hotter than conventional nuclear plants. That extreme temperature enables the sulfur-iodine cycle for efficient hydrogen production and direct industrial heat applications.

The safety profile relies on physics, not operators. TRISO fuel particles are uranium kernels wrapped in silicon carbide layers that can't melt down. If cooling systems fail, the reactor shuts itself down through passive safety mechanisms. Taylor made the controversial claim that spent fuel from their reactors could be held in human hands with radiation exposure equivalent to a CT scan. Nuclear engineers called this an exaggeration, but it signals Valar's confidence in their design.

The reactors are containerized and transportable by truck. Ward One, their first nuclear reactor named after Taylor's great-grandfather, took about 10 months to build. The company emphasizes speed and replication over custom engineering. Each reactor is designed to be identical, enabling mass manufacturing that drives down per-unit costs and deployment timelines.

This dual-use capability extends beyond electricity generation. The reactors can power data centers, forward military bases, hydrogen production facilities, and synthetic fuel plants. High-temperature nuclear heat combined with captured CO₂ can produce carbon-neutral diesel, jet fuel, and gasoline. Each installation becomes a self-contained energy ecosystem operating independently of the electrical grid.

Market Strategy & Business Model

Valar targets customers who need massive, continuous power but can't rely on grid connections. Defense installations, heavy industry, and AI data centers top the list. Military bases need reliable electricity without vulnerable diesel supply lines. Steel mills and aluminum smelters require enormous amounts of process heat. Data centers for AI face power shortages as demand explodes.

The breakthrough strategy involves clustering hundreds of reactors at single "gigasites." This approach spreads regulatory and site preparation costs across multiple units while achieving gigawatt-scale output. Think Tesla's Gigafactories, but for nuclear power.

Valar signed a contract with the Philippines Nuclear Research Institute to build Ward One overseas. The Philippines provides a more favorable regulatory environment than the United States, allowing Valar to demonstrate their technology before attempting domestic deployment. The company also announced a partnership with Utah in mid-2025 to explore building a prototype reactor in the state.

Long-term customers likely include the Department of Defense, which is actively pursuing microreactors for bases, and industrial companies seeking energy independence. AI infrastructure builders and synthetic fuel producers represent growing markets as these sectors scale rapidly.

Funding & Regulatory Tactics

Valar raised $19 million in seed funding in February 2025, led by Riot Ventures with participation from AlleyCorp, Initialized Capital, Day One Ventures, and Steel Atlas. The company has raised approximately $21 million total, including earlier pre-seed funding. No major federal grants have been disclosed, suggesting Valar prioritizes speed and flexibility over government funding constraints.

The regulatory strategy involves direct confrontation with the Nuclear Regulatory Commission. In April 2025, Valar joined a multi-state lawsuit against the NRC alongside Texas, Utah, Florida, Louisiana, Arizona, and fellow nuclear startups Last Energy and Deep Fission. The lawsuit argues that small, inherently safe reactors shouldn't require the same licensing gauntlet as gigawatt-scale designs.

This legal challenge represents a bet that nuclear regulation will evolve to match technological reality. Traditional reactor licensing takes over a decade and costs hundreds of millions of dollars. Valar argues their reactors are fundamentally different and deserve streamlined oversight.

The company remains dependent on Department of Energy efforts to establish domestic HALEU fuel production. High-Assay Low-Enriched Uranium is required for advanced reactors but currently in limited supply. Most HALEU historically came from Russia, creating supply chain vulnerabilities as those imports face restrictions.

Strategic Relevance

Grid independence transforms how critical infrastructure operates. Data centers, hospitals, and semiconductor fabs can't afford power outages. Microreactors provide localized, always-on energy that removes dependency on fragile grid infrastructure. Military bases get long-duration power without diesel convoys that create logistical vulnerabilities.

Industrial reshoring becomes viable when energy is portable. High-temperature nuclear heat provides the missing link for domestic heavy manufacturing. Steel production, aluminum smelting, and chemical processing can locate near raw materials or ports rather than existing grid infrastructure. This supports both Biden and Trump administration initiatives to reshore strategic industries.

The defense implications extend beyond base power. Aligns with Project Pele and broader DoD interest in mobile reactors for contested environments. Forward bases in the Pacific or Arctic could maintain full operations with nuclear power, enhancing deterrence through energy sovereignty. Distributed nuclear infrastructure makes the national energy system more resilient to cyberattacks or natural disasters.

Synthetic fuel production adds another strategic dimension. Nuclear-powered facilities can convert atmospheric CO₂ and water into jet fuel, diesel, and gasoline. This capability reduces petroleum import dependence while providing carbon-neutral alternatives to fossil fuels. Each gigasite becomes a domestic refinery powered by uranium instead of crude oil.

Challenges & Risks

NRC resistance remains formidable even if the lawsuit succeeds. Nuclear regulation exists for valid safety reasons, and public trust requires demonstrated performance over time. The path to widespread domestic deployment faces years of bureaucratic challenges regardless of legal victories.

HALEU fuel supply represents a critical bottleneck. Department of Energy plans to establish domestic enrichment capacity remain underfunded and behind schedule. Competition for limited HALEU supplies intensifies as more advanced reactor companies progress toward deployment.

Public skepticism of nuclear power persists, especially regarding clustered reactor installations. The gigasite model concentrates nuclear infrastructure in ways that may trigger local opposition. Safety communications require precision, and Taylor's claims about handling spent fuel drew criticism from nuclear engineers as potentially misleading.

Technical execution at scale remains unproven. The gigasite business model depends on regulatory costs being spread across hundreds of reactors, an assumption untested in practice. Supply chain challenges for specialized nuclear components could limit manufacturing scalability. Any technical setbacks or safety incidents would significantly impact public and investor confidence.

Tactical power means more than battlefield energy. It means industrial sovereignty, grid independence, and the ability to build where resources exist rather than where utilities allow. Valar wants to make kilowatts as mobile and scalable as trucks, transforming nuclear from a centralized utility into distributed infrastructure.

Success here breaks nuclear technology out of its Cold War legacy mindset. Failure reinforces the perception that nuclear can't move fast enough to matter in the AI era. Either way, Valar Atomics forces a fundamental question: can nuclear build at the speed of progress?

Join Forge Ahead for weekly news and actionable insight on solving America’s biggest problems.

Subscribe now

]]><![CDATA[Spotlight: Shield AI]]>https://lukeringlein.substack.com/p/spotlight-shield-aihttps://lukeringlein.substack.com/p/spotlight-shield-aiSun, 27 Jul 2025 20:12:13 GMT

Stay sharp on the companies shaping the next generation of defense. Subscribe to Forge Ahead for more spotlights like this—and see who’s building the tools that will define tomorrow’s wars.

Subscribe now

As far as American Dynamism start ups go, you can’t get much cooler than Shield AI.

Vulture droid from Star Wars.

Think Ender’s Game, Star Wars, HAL 9000— take your pick. The concept of AI powered planes/drones/spaceships is nothing new.

Shield AI has finally taken that concept and brought it out of science fiction.

They're developing autonomous air systems by creating new aircraft like the Nova and V-BAT and integrating their AI brain into existing ones, like the MQ-20 Avenger.

Most impressivle, Shield has scaled ruthlessly. After only ten years since founding, their systems have already flown hundreds of combat missions. The V-BAT has flown 130 missions in Ukraine, consistently navigating heavily jammed airspace and penetrating Russian defenses as far as 100km behind enemy lines.

To date, this performance has made significantly more impact than even Anduril’s systems.

Who They Are

Shield was founded in 2015 by brothers Ryan and Brandon Tseng. Ryan is an ex-Silicon valley exec, and Brandon a former Navy SEAL. They’ve built a team marrying entrepreneurial expertise with operational experience. Their mission is “to protect service members and civilians with intelligent systems.”

tldr; Shield isn’t just building drones, they’re rewriting what airpower looks like in the AI era.

What They Do: AI Pilots, Not Just Drones

On the surface, Shield appears to be primarily an aircraft manufacturer. In fact, their core product is their AI backend: Hivemind. Officially, hivemind is an aircraft-agnostic “AI-pilot”. It integrates into unmanned systems and allows them to navigate them without communication or GPS in contested/jammed environments.

This system can be ‘downloaded’ (gross simplification; forgive me, tech bros) into existing aerial systems and trained to pilot them. This has been done with a few major aircraft, most notably General Atomic’s MQ-20 avenger drone.

Detailed view
DVIDS

Importantly, Shield has designed Hivemind to integrate quickly. Their last deployment took only 165 days from contract to first flight. Relative to typical defense timelines, that’s breakneck speed.

These systems are currently designed to conduct ISR (Intelligence, Surveillance, Reconnaissance), strike missions, electronic warfare, and support manned-unmanned teaming. They sell to the US and allied defense forces, notably Ukraine and Australia.

Most importantly, Shield builds AI-first, modular systems that are designed to scale. They can upgrade software weekly, vs. legacy systems that take decades to iterate. As of spring 2024, it had flown 6 unique aircraft, including three types of quadcopters, the V-BAT VTOL drone, a Kratos Firejet, an F-16, and it continues to add more.

Like Anduril, they build first, walk into the DoD with a working product and a pitch, and sell. This has allowed them to scale rapidly and beat out the contract-eating primes.

III. How They’re Doing It: Speed + Silicon Valley DNA

Shield’s edge comes from pairing Silicon Valley’s fail-fast culture with military-grade discipline. Hivemind is trained like a self-driving car brain on steroids. It uses reinforcement learning and high-fidelity simulation to run millions of virtual combat hours before touching real hardware.

By the time a drone or jet runs its first mission, the AI pilot has “flown” more than most fighter squadrons ever will. This mirrors the approach of companies like Waymo, except Shield is tuning for chaotic battlefields instead of suburban traffic.

The funding story reads like a who’s who of American venture capital. Shield has raised over $600 million from heavy hitters like Andreessen Horowitz and Founders Fund and is valued at $5.3 billion as of March 2025. That makes them the second biggest defense tech start-up, trailing only Anduril.

For a defense startup, that’s unicorn territory usually reserved for software giants. This money is fueling an aggressive R&D pipeline, with hardware and software teams sharing the same floor to slash iteration cycles.

Partnerships keep pushing the tech forward. Shield is working with the U.S. Navy and Air Force on manned-unmanned teaming, where Hivemind drones act as wingmen for F-35s and other fighters. Unlike many defense startups that get lost in endless prototypes, Shield has already proven its systems in real-world combat, from 130+ V-BAT missions in Ukraine to AI-controlled F-16 flights during DARPA’s ACE program.

Their mantra is simple: build it, fly it, break it, fix it, repeat. That relentless pace has made them the most combat-tested AI autonomy company in the defense tech sector.

“Readiness means flying today, not waiting for perfect conditions.

Autonomy doesn’t get better in a lab or inside a PowerPoint. It improves when it flies, when it fails, adapts, and flies again.”

—Christian Gutierrez, VP of Hivemind Solutions at Shield AI

Why it Matters: Strategic Context

The US and China are in an arms race to weaponize AI and integrate autonomy into their systems. The first nation to successfully optimize will have a significant advantage: these systems are cheaper, safer, and more scalable. There will never be a shortage of AI pilots (that is, if we can source enough chips to power them), making them expendable. Even better, its cheap! For the price of a singular MQ-9A Reaper, we could deploy dozens of V-BATs. Swarm technology is rapidly developing; both the US and the PLA have made significant investments into harnessing it.

The Pentagon is already planning for this; one deployment strategy envisions a drone ‘hellscape’ to defend Taiwan in the event of invasion. And after observing Ukraine’s innovative drone use, militaries around the world are preparing for the next war— one where small, cheap drones dominate.

Defense startups like Shield AI can deploy AI powered capabilities faster and at lower cost per capability than traditional primes.

In June 2025, Shield AI and AWS launched a cloud‑based firmware pipeline that lets engineers push secure, battle‑ready software updates over‑the‑air to deployed Hivemind systems. This allows for streamlined and adaptive software updates within days, not weeks.

V. Challenges and Criticisms

Unfortunately, Shield still has to deal with the same inefficient procurement systems as everyone else. They will need to find a way to scale without being curbed by that bottleneck. A 2023 House Armed Services hearing concluded that AI evolves far faster than DoD acquisition timelines, often rendering new tools outdated by the time they reach the field. If the US wishes to keep a technological advantage over its adversaries, we will need to find a way to streamline procurement for bleeding-edge tech.

At the same time, critics raise concerns around fully autonomous combat systems. Certainly, removing humans completely from the kill chain would be a mistake, but that doesn’t mean we can’t move forward with autonomy while remaining moral.

Instead of overregulating autonomous weapons, we need to find a way to add safeguards without substantially slowing development. The truth is, autonomous weapons aren’t going away. There won’t always be a human pulling every trigger. But every delay in procurement is a gift to an adversary who doesn’t wait.

Shield AI assures us that its AI pilots will always have ‘human intent’ guiding them to retain human-on-the-loop control.

VI. The Future of Shield AI

Shield AI’s roadmap is aggressive: the company is already moving beyond drones, integrating Hivemind into larger aircraft like the F-16 and next-generation UCAVs. It flew a General Atomics MQ-20 Avenger autonomously, a clear signal that the same software could one day run loyal wingmen alongside F-35s or even fully unmanned fighter jets.

On the international front, allied nations are lining up. The global military UAV market is projected to grow from $36 billion in 2023 to $88 billion by 2030, and countries from India to the UK are investing heavily in AI-powered drones. Shield AI’s partnerships with the U.S. Navy, Air Force, and foreign allies position it to capture a meaningful share of this demand.

Brandon Tseng has made the company’s vision plain: every U.S. aircraft, from MQ-9s to stealth bombers, should have an AI co-pilot before the decade closes. Christopher Kubasik CEO of L3Harris insists that Shield AI is proving that “autonomy at scale” isn’t just possible; it’s inevitable.

If they succeed, it could redefine the defense ecosystem itself. The traditional primes move too slowly to keep up with software-defined warfare. Shield is proving that small, fast-moving teams can build the brains of tomorrow’s military power.

If the next war is fought at machine speed, the winner will be whoever can teach machines to fight for them first. Shield AI wants that edge.

Stay sharp on the companies shaping the next generation of defense. Subscribe to Forge Ahead for more spotlights like this—and see who’s building the tools that will define tomorrow’s wars.

Subscribe now

]]><![CDATA[Spotlight: Hadrian]]>https://lukeringlein.substack.com/p/spotlight-hadrianhttps://lukeringlein.substack.com/p/spotlight-hadrianSun, 20 Jul 2025 21:49:03 GMTSection I: Introduction

American factories rarely make headlines unless something goes wrong. For decades, the aerospace and defense supply chain has leaned on small, slow-moving machine shops where lead times stretch for months and skilled machinists are aging out of the trade. Hadrian is betting that software, automation, and a fresh approach to factory design can break that cycle.

The company isn’t chasing hype. It’s solving a clear problem: the steady decline of U.S. manufacturing capacity. “America cannot afford to lose another generation of industrial capacity,” said Chris Power, Hadrian’s founder and CEO (TechCrunch, 2024). Launched in 2020, Hadrian focuses on producing critical aerospace and defense components faster and with fewer costs, all while keeping the work onshore.

Hadrian is part of a growing set of companies proving that modern factories can be efficient and competitive. For anyone following the return of American industry, this is a company worth watching.

Manufacturing startup Hadrian to expand to Arizona, and into defense primes' own factories - Breaking Defense
Hadrian factory floor

Section II: Who Is Hadrian?

Hadrian is a manufacturing startup founded in 2020 by Chris Power, an Australian who moved to the U.S. with just a few thousand dollars and a belief that America could still lead in making things. Based in Southern California, the company builds highly automated factories that produce precision parts for aerospace and defense.

Power’s path wasn’t typical. He spent months cold-calling factories across Texas, trying to find a problem worth solving. Aerospace and defense stood out: too many critical components were made in outdated shops with slow turnaround times and thin margins. He saw a chance to combine robotics, artificial intelligence, and proprietary software to modernize the process.

From day one, Hadrian’s identity has been tied to this mission. The company is staffed with engineers and operators who see themselves as rebuilding industrial capacity, not just making parts. As one of the standouts in the “American Dynamism” movement, Hadrian has drawn interest from investors and government agencies alike (TechCrunch, 2025).

Section III: What Does Hadrian Do?

Hadrian builds high-precision metal parts for rockets, satellites, jets, and other aerospace and defense systems. Its factories run advanced CNC machines, guided by software that automates programming and scheduling. This setup allows a single operator to manage several machines at once, reducing both time and cost (TechCrunch, 2025).

The company started with CNC machining but has expanded into welding, casting, and 3D printing to produce a wider range of components. By integrating these capabilities under one roof, Hadrian can deliver parts in weeks instead of months. The goal is to eliminate bottlenecks that slow down the aerospace supply chain.

Hadrian’s approach is not just about speed. It’s about reliability. Traditional job shops often struggle with backlogs and labor shortages, but Hadrian’s automated workflow is designed to run at scale without relying on a handful of expert machinists.

Section IV: When and How They Rose

Chris Power founded Hadrian in late 2020, just a year after arriving in the U.S. with $6,000 and a conviction that American manufacturing could be rebuilt. The company quickly raised a $9.5 million seed round and built its first 20,000-square-foot R&D factory in Hawthorne, California (TechCrunch, 2024).

Growth came fast. By early 2022, Hadrian secured $90 million in Series A funding and opened a second, much larger facility in Torrance. Just a year later, the company had over $20 million in revenue and a backlog of orders from major aerospace and defense players. The momentum attracted another wave of investors, bringing total funding to over $350 million by 2025 (Breaking Defense, 2025).

Hadrian’s rise is built on execution, not just capital. The company developed a playbook for rapidly standing up new factories and refining its automation platform. In 2025, it launched a third site in Mesa, Arizona, with plans for four or five more facilities within the year.

Defense manufacturing startup to launch 'AI-powered' factory in Mesa next year

Section V: Why Does Hadrian Exist?

Hadrian was built to solve a quiet crisis: America’s industrial capacity is thinning out. Too many critical parts for aerospace and defense are still made in small shops with aging machines and a shrinking workforce. Chris Power saw this as more than an economic problem. He viewed it as a national risk.

“When great powers outsource their ability to build, they don’t stay great for long,” Power has said (TechCrunch, 2024). He believes that relying on fragile supply chains leaves the U.S. vulnerable in a future conflict or space race. For him, policy talk isn’t enough. What matters is building capacity onshore, at scale, and at speed.

Hadrian exists to make that vision real. By bringing automation to advanced manufacturing, the company wants to ensure the U.S. can produce the hardware it needs—rockets, satellites, weapons—without depending on foreign suppliers. It’s not just about machines; it’s about keeping industrial know-how alive for the next generation.

Section VI: How Does Hadrian Operate?

Hadrian treats its factories like products. The company built a proprietary software platform, Opus, that handles almost every step of production—from generating CNC instructions to scheduling jobs and managing quality checks (PR Newswire, 2025). Opus automates 80–90% of the work, leaving operators to focus on the remaining tasks that require human judgment.

This model changes the role of the worker. A single technician can now oversee four to six machines, instead of babysitting just one. It’s not about replacing skilled machinists but extending their capabilities and building a workforce that doesn’t rely on decades of “tribal knowledge.” In its Torrance facility, Hadrian reports productivity that’s three to four times the industry average (TechCrunch, 2025).

Hadrian also offers a “Factories-as-a-Service” model. Defense primes can outsource entire programs to Hadrian, which either builds a dedicated facility or embeds its automation systems into existing ones. To scale fast, Hadrian has built a cookie-cutter playbook for launching new sites—rolling out a new facility in under six months. In 2025 alone, they plan four to five additional factories to meet defense demand (Breaking Defense, 2025).

Section VII: Mission and Its Impact

Hadrian’s mission is simple but ambitious: reindustrialize America. The company wants to make sure that advanced technology, especially for aerospace and defense, can be built domestically at competitive cost and speed (PR Newswire, 2025).

The impact is already visible. Faster part production means rockets, satellites, and precision weapons can move from design to field deployment without months of delay. For the Department of Defense and private aerospace companies, that kind of turnaround is critical. It’s not just about making parts; it’s about making sure the U.S. can respond quickly when stakes are high.

Hadrian also affects the workforce. By automating the grunt work and training operators on modern systems, it’s helping to attract younger talent to manufacturing. “We’re supercharging the American worker,” Power said, noting that automation isn’t a replacement but a force multiplier for skilled labor (Breaking Defense, 2025).

Section VIII: Lessons and Takeaways

Hadrian’s rise offers lessons for anyone who cares about building. The first is that a clear mission attracts more than capital. Founders Fund and a16z didn’t just back Hadrian because of its tech; they saw a company taking on a strategic challenge that matters for the country (TechCrunch, 2025).

The second lesson is that “old” industries aren’t dead, they’re just under-innovated. Precision machining hasn’t changed much in decades, but Hadrian used robotics and software to boost productivity by three to four times. That kind of efficiency shift can happen anywhere traditional operations have been left to stagnate.

Third, automation doesn’t have to kill jobs. Hadrian’s model shows how you can augment workers instead of replacing them, preserving essential expertise while bringing new talent into the fold. It’s a blueprint for any company looking to modernize without hollowing out its workforce.

Scaling hardware like software is another takeaway. Hadrian replicates factories with a playbook as standardized as a fast-food franchise rollout. Policymakers could take note: speed matters, and red tape slows more than just permits—it slows national strength.

Finally, Hadrian is part of a bigger picture. It’s a reminder that American dynamism isn’t just about new apps or AI models. It’s about companies willing to solve unglamorous problems—like who’s going to make the parts for the next generation of rockets.

Hadrian raises $260M to build out ...

Section IX: Conclusion

Hadrian is proving that advanced manufacturing can move as fast as tech startups. Its factories aren’t showpieces, they’re workhorses designed to cut lead times, rebuild supply chains, and give the U.S. a competitive edge in space and defense. For a country that has too often outsourced its backbone, this matters.

The company’s next challenge is scaling without losing its edge. If Hadrian can replicate its model across the country, it could become a cornerstone of the new industrial era. This is worth watching—not just for aerospace insiders, but for anyone who believes America should be able to build what it needs, when it needs it.

If you care about rebuilding America’s capacity to make things that matter, follow companies like Hadrian. They’re not just fixing supply chains; they’re shaping the future of national strength.

Follow along

]]><![CDATA[Spotlight: Who is Anduril, Really?]]>https://lukeringlein.substack.com/p/spotlight-who-is-anduril-reallyhttps://lukeringlein.substack.com/p/spotlight-who-is-anduril-reallySun, 13 Jul 2025 16:06:24 GMT

Welcome to the first installment of Sunday Spotlight, where we zoom in on one company working ‘toward the national interest’. We explore the who, what, why, and how through an unbiased lens.

Subscribe so you don’t miss next week’s article about Hadrian.

Palmer Luckey got rich building headsets to escape reality. Now, he’s building weapons to protect it.

In 2017, he was 24, flush with Oculus money, and best known for inventing VR goggles in his garage. Then he quietly donated to a group that ran pro-Trump memes. He wasn’t giving speeches or picking fights, but he was politically active in a city that rewards silence. Facebook forced him out.

Before you cringe, this article isn’t about politics. It’s about what Luckey did after.

Most people assumed he'd disappear into a private island or a hardware incubator. He chose to build missiles.

Luckey looked at the Pentagon and saw a factory that couldn’t ship. Tools that didn’t talk to each other. Software duct-taped to hardware. Contractors paid more to be late than to deliver. Meanwhile, China was fielding drones, scaling defense startups, and wiring its military with AI. America was holding hearings.

He decided to build a company that could move faster than the primes and think sharper than the generals. One that didn’t wait to be asked. If the Pentagon needed a product, it would already exist. If it didn’t bite, too bad. Someone else would.

At the same time, Big Tech was pulling back. After Google employees revolted over the company’s work on a drone AI contract, leadership folded. That wasn’t just a PR retreat—it was an industrial abdication. The people with the talent to defend the country had decided to sit the fight out.

That was the gap. He stepped into it.

Subscribe now

Section II: Building Anduril

Palmer Luckey had the vision. But the firepower came from a crew of engineers and operators who’d already tried to fix defense from the inside—and failed.

The founding team included Trae Stephens, a former Palantir exec and partner at Founders Fund, along with Matt Grimm and Brian Schimpf, both from Palantir’s early days, and Joe Chen, a longtime software lead. They’d seen how broken the process was. The Pentagon didn’t lack money. It lacked delivery. Getting code into production took years. Hardware came slower.

In 2017, they launched Anduril Industries. Named after the Lord of The Rings sword “The Flame of the West", the goal was to flip the model. Build fast, fund it privately, and let the product speak. Stephens made it plain: traditional defense firms didn’t understand software, and Silicon Valley didn’t want the responsibility of national defense.

Their first test came fast. Anduril’s Sentry Tower, a solar-powered surveillance post with radar and thermal cameras, was deployed in a 10-week field trial in Texas. Instead of streaming footage to a human operator, it used Anduril’s software to autonomously flag movement. The towers detected 55 people and 982 pounds of marijuana, with minimal operator input. Border Patrol confirmed the system outperformed expectations.

Anduril Launches Extended Range Sentry Tower (XRST) to Advance Counter Intrusion Missions | Anduril
The tower that got them in the door. Built in-house. Deployed in months.

That pilot turned into a real contract. By 2020, DHS had awarded Anduril a deal to install 60 towers across the U.S.-Mexico border. Each unit cost under $500,000—less than a fifth of what Boeing had charged for a failed attempt years prior. Anduril wasn’t just cheaper. It worked.

The towers weren’t revolutionary by themselves. The process was. Build fast. Deploy early. Iterate on real-world feedback. And do it all without waiting for a contract. The startup hadn’t just broken in—it had shown up the old guard.

Section III: The Operating System of War

Anduril never wanted to be just a drone company. The towers were proof of concept. The real product was the software.

That software is called Lattice. It’s not a platform, not a dashboard, not a buzzword. It’s a command-and-control operating system that takes inputs from radars, drones, towers, and cameras, then fuses them into a real-time picture of the battlespace. It flags threats automatically and suggests what to do next. You can run it on a laptop in a warzone. You can run it in VR if you want to get fancy.

Anduril used Lattice in a 2020 simulation with the Air Force to track incoming cruise missiles. Operators saw threats rendered in a 3D battlespace—live, not lagging. Then they simulated a response, using the same system that would be deployed in the field.

Everything else Anduril builds plugs into Lattice. That’s how they move faster than legacy firms. They don’t just bolt software onto a weapon system after it’s built. The software comes first.

The clearest example is Ghost—a tactical drone built to fly autonomously and carry serious payload. It’s light enough to launch from a backpack. It’s quiet. It can follow targets using thermal vision and onboard AI. The Ghost-4 model can fly for 100 minutes and carry 35 pounds of gear. You don’t pilot it. You task it.

Then there’s Anvil—a quadcopter designed to intercept and smash enemy drones out of the sky. The latest version, Anvil-M, adds a detonation system. It can be deployed as part of a layered drone defense net, with Lattice triggering the intercept the moment a threat appears. That’s not theory. The Marine Corps awarded Anduril a $642 million contract to field this counter-drone capability across its forward bases.

Next comes Roadrunner. Imagine a cruise missile that can be reused. Roadrunner is a twin-jet autonomous interceptor that can take off vertically, strike, or land itself and get reloaded. It's modular. The same airframe can carry ISR sensors or warheads. It’s not a concept sketch; it’s in low-rate production now.

Underwater, Anduril is deploying Dive-LD, an extra-large autonomous undersea drone. In Australia, it’s branded as Ghost Shark, built in partnership with the Royal Navy under a $70 million co-development agreement. It can launch a swarm of smaller Copperhead drones from its payload bay. Defense One reports it’s already completed successful prototype runs.

Anduril demonstrates multi-domain unmanned systems command and control, undersea, and airborne capabilities at US Navy's IBP 24.1 | Anduril
Lattice connects Anduril’s systems, from towers to drones to missiles,into one decision layer.”

These aren’t just cool gadgets. Together, they form a kill chain. Sensors detect. Lattice decides. Drones act. Humans stay in the loop, but the system creates a network that can handle the full workflow at machine speed. Anduril isn’t building a family of weapons. It’s building the nervous system that connects them.

Section IV: The Anduril Way

Anduril doesn’t chase contracts. It builds first and sells later. That one shift flips the entire defense playbook.

Legacy defense firms wait for the Pentagon to issue a formal request, then spend years writing proposals, lobbying program officers, and spreading risk across bloated timelines. Anduril skips that. It uses private capital to prototype weapons up front, then walks into the room with working gear.

This is what they mean when they talk about “product-led defense.” No PowerPoints. No studies. Just systems that can be bought today.

Anduril’s business model also makes it hard to kill. Instead of relying on cost-plus development, it builds fast, funds it internally, and gets paid when customers see results. As Defense One reported, the company has built a “private pipeline of autonomy-enabled systems” that let it bypass the usual bureaucratic churn.

That model is scaling. In 2023, Anduril won a 10-year, $642 million contract with the Marine Corps to provide a layered counter-drone system based on Lattice, Ghost, and Anvil. It’s also been selected as one of two vendors—alongside General Atomics—for the Air Force’s Collaborative Combat Aircraft (CCA) program, beating out giants like Boeing and Lockheed.

Anduril didn’t stop there. When Microsoft’s augmented reality headset for the Army (IVAS) ran into delays, Anduril was tapped to rescue the program. The original contract was worth up to $21.9 billion.

These aren’t niche wins. They’re flagship programs. Anduril’s systems are now inside frontline units, forward operating bases, and special operations kits. Fortune reported that the company’s revenue doubled to roughly $1 billion in 2024. It raised a $2.5 billion Series G, pushing its valuation to $30.5 billion.

Anduril isn't trying to be a contractor. It's trying to become the next prime—without inheriting the bloat. The bet is simple: whoever builds fastest, wins.

Section V: Critics, Watchdogs, and Warnings

Not everyone sees Anduril as a national asset. Some see it as a warning sign.

The company’s first big contract—deploying autonomous surveillance towers at the southern border—sparked immediate backlash. Civil rights groups like Mijente and the Center for Media Justice argued that automated detection systems pushed migrants into more dangerous terrain and normalized surveillance in remote regions. The ACLU called it a “dangerous expansion of surveillance infrastructure.”

Anduril weaponized drone factory protest

Anduril pushed back. The towers don’t use facial recognition. They don’t classify by race, gender, or identity. They detect objects and track motion. That’s it. But the bigger fear isn’t just how they work today, it’s how easily they could scale beyond border zones. From deserts to cities. From national security to domestic policing.

Then there’s the autonomy problem. The company says humans always stay in the loop when lethal force is on the table. But the whole point of Lattice is speed. If your kill chain runs faster than a person can think, are they really in control? Critics worry about that slippery slope, especially as the U.S. faces pressure to match China's gains in battlefield AI.

But here’s the truth behind the noise: the people most upset about Anduril aren’t worried it doesn’t work. They’re worried it works too well, and changes the way wars get fought.

Section VI: Forward Edge

Anduril isn’t just scaling. It’s expanding into every domain the Pentagon says it wants to dominate—air, sea, land, space, and now manufacturing.

Anduril’s newest play? Manufacturing itself. In 2023, it unveiled Arsenal, a vertically integrated smart factory designed to produce drones, missiles, and sensors with short-cycle iteration. Think SpaceX’s rocket shop, but for autonomous defense systems. The goal isn’t just volume. It’s control. “We can’t afford to separate design from production anymore,” Luckey said at the launch.

Anduril is also plugged into Replicator, the Pentagon’s fast-track initiative to mass-produce thousands of low-cost autonomous systems. The whole point is to beat China’s scale with software and speed. Anduril is already supplying core platforms.

If the primes are trying to protect programs, Anduril is trying to replace them. What it’s really building isn’t just a portfolio. It’s a blueprint for how to outpace an adversary without outspending them.

$2B Arsenal-1 Megaproject: OC Firms Join Forces for Anduril's Massive Ohio Site - Orange County Business Journal
Anduril’s Arsenal factory gives it end-to-end control over design, build, and deployment timelines.

Section VII: No One’s Coming

Anduril doesn’t pretend to fix the system. It works around it.

That’s the bet: build fast, fund it yourself, and deliver before the bureaucracy even starts writing requirements. Whether that model scales, or gets buried under the weight of the system it’s replacing, is still unclear.

But if war comes fast, Anduril’s already moving.

The executives behind Palmer Luckey's new digital banking startup Erebor include banking, Big Law, and tech vets

Subscribe so you won’t miss next week’s installment of Sunday Spotlight: Hadrian

Sources/Further Reading

]]>