A $6,000 Quadcopter Is Outshooting $13 Million Missiles Over Kyiv.

The Smart-Glasses Founders Betting That Air Defense Can Be Mass-Produced

The 80mm Seeker That Could Change Air Defense Economics

An 80-millimeter radar seeker that costs four thousand dollars just flew on a Ukrainian interceptor drone. Two years ago that sentence would have sounded absurd. The seeker head has always been the single most expensive subsystem in an autonomous interceptor: six figures, years of development, hundreds of engineers at a major defense contractor. Vernius Systems, a San Francisco startup spun out of a volunteer nonprofit called Defense Tech for Ukraine, built this one in weeks. They demonstrated it in August on the STING interceptor, a $2,000 FPV drone developed by the Ukrainian manufacturer Wild Hornets. The STING has already claimed several thousand Shahed and Gerbera kills since its first confirmed hit went viral in spring 2025. Backing includes former National Security Advisor LTG H.R. McMaster, UA1, Pioneer Fund, Rebel Fund, and Palm Drive Capital, plus a Y Combinator Summer 2026 slot and a $125,000 micro-seed.

The economics of air defense have inverted. A Shahed-136 costs Russia perhaps to produce. Ukraine's best option for shooting one down without a Patriot missile — Vernius reports $4M per Patriot intercept. The single most expensive subsystem of every autonomous interceptor is its radar seeker head. — is a pilot flying an FPV drone by eye. That works in clear weather. It fails in fog, smoke, rain, or darkness. Serhii "Flash" Beskrestnov, now an advisor to President Zelenskyy, wrote last month that winter weather can cut visibility to a few dozen meters and send interceptor success rates off a cliff. More than half the Shaheds over Ukraine are now jet-powered variants — Geran-3, Geran-4, Geran-5 — flying too fast for manual pilots to catch reliably. Manual interception currently yields a 20 to 50 percent kill probability, Vernius says, and that number drops further when the pilot cannot see the target.

The Archimedes 1A changes the sensor economics. It operates at 24 GHz with a fixed beam, a 25 by 14 degree field of view, and a claimed detection range beyond one kilometer against a Shahed-class radar cross section of roughly 0 dBsm. The unit weighs little enough to bolt onto an existing interceptor airframe. It runs two modes. In fully autonomous interception, the interceptor receives a cue from a command-and-control network (bearing, range, altitude), uses the radar for midcourse guidance beyond visual range, then hands off to autonomous vision-based terminal guidance for the final approach. In operator-assisted mode, the seeker feeds live target range and bearing to the pilot, letting them acquire the target before it becomes visible. The same hardware does both.

Vernius says the seeker is designed to adapt to any existing interceptor platform. The company has also listed integration with jet-powered interceptor UAVs including the Alexa Spatium system, which would extend the engagement envelope against the faster Geran variants. The STING demonstration matters because Wild Hornets' interceptor is already operational at scale. Ukraine's defense ministry reported more than 33,000 enemy drones destroyed by interceptors in March 2026 alone, double the previous month. Adding a $4,000 radar seeker to a $2,000 airframe creates a $6,000 autonomous interceptor — roughly the price of a high-end bicycle — that can hunt a target costing five to ten times as much in weather that blinds every other cheap option.

The seeker is not a phased array. It does not steer its beam electronically. That simplicity is the point. Vernius built it on a printed-circuit-board architecture that can be fabricated in standard commercial shops, not a defense-unique production line. The company's founders, Arthur Nguyen-Cao, Hisham El-Halabi, and Joseph Gonzalez, bring a decade of RF PCB engineering across those programs and FPV hardware already fielded by frontline Ukrainian units. They are not trying to match the performance of a Raytheon seeker. They are trying to undercut its cost by two orders of magnitude while delivering enough capability to make autonomous interception work in the conditions where Ukraine actually fights.

If the Archimedes 1A performs in combat the way it has in demonstration, the interceptor drone stops being a clever workaround and becomes something the Pentagon has spent decades and billions trying to field: a surface-to-air missile that costs less than the target it kills.

How Smart-Glass Engineers Became Missile Makers

Arthur Nguyen-Cao and Hisham El-Halabi spent years shrinking compute, sensors, and batteries into a frame that sits on a nose. Their previous company, Auctify, launched the Specs, AI smart glasses that tracked eye movement, detected procrastination, and promised productivity gains, on Indiegogo in September 2020 after raising $250,000. The Verge covered the launch. The Center for Data Innovation interviewed the co-founders about motivation engineering. By 2025 the smart-glasses category had consolidated around Meta and EssilorLuxottica, which together hold more than 80 percent market share, and Meta was preparing a $299 consumer line for June 2026. Auctify had already pivoted toward a calorie-tracking variant. The consumer wearable window was closing.

The hardware lessons did not evaporate. Building Specs forced the team through the same gauntlet that a missile seeker demands: multi-layer RF PCB layout in a volume the size of a thumb, power budgets measured in milliwatts, thermal management without a fan, and a ship date that could not slip. Nguyen-Cao describes the overlap directly: "Between us we have over a decade of hardware experience, and we're building the seeker that's going on tens of thousands of interceptors a month." El-Halabi, who studied engineering science at the University of Toronto, brought the same RF and signal-integrity discipline to both products. Joseph Gonzalez, the third co-founder and COO, rounds out a team that also includes those programs and FPV drone hardware already flown by Ukrainian frontline units.

The pivot from consumer wearable to defense interceptor is not a one-off. Y Combinator accepted Vernius into its Summer 2026 batch and wrote a $125,000 micro-seed check in June. The same batch structure that once funneled founders into SaaS and fintech now routes them toward radar seekers, solid-rocket motors, and autonomous surface vessels.

The job posting for Vernius's first engineering hire makes the talent profile explicit: RF PCB design experience, at least one ambitious self-directed board shipped on an aggressive timeline, US citizenship, with bonuses for phased-array PCB experience, prior hardtech founding experience, and published research. The requirements read like a filter for engineers who have already survived a hardware tape-out under pressure. In the first 30 days the hire finalizes the V2 interceptor radar seeker layout; by day 90 they oversee the first 1,000 boards through assembly. The boards ship on over 10,000 interceptors monthly starting in 2027.

That tempo — 30 days to layout, 90 to production — mirrors the founders' own trajectory. Nguyen-Cao led 14 engineers at Defense Tech for Ukraine, shipping FPV hardware to frontline units on timelines measured in weeks. El-Halabi put a satellite in low Earth orbit. Gonzalez architected secure communications for the Navy's Tomahawk fleet under Q clearance at Sandia National Labs. All three co-founded Auctify. They know what aggressive hardware schedules look like because they have lived them.

"Developing a radar seeker takes a 200+ person team at Raytheon and hundreds of millions of dollars," the company states on its YC profile. Vernius is three people.

The posting also reveals where the tooling is headed. "Build our in-house AI tooling for PCB design, with the long-term goal of a pipeline that generates and validates boards with very little human time." Vernius is betting it can compress the radar seeker development cycle further by writing its own automation layer atop existing EDA tools.

Industry data confirms the hiring friction. Radar systems engineers average annually, and the specialized skill set (radar physics, RF layout, signal integrity, embedded firmware) makes qualified candidates scarce. Vernius's approach: skip the generalist radar engineer and hire a PCB specialist who has already shipped boards in anger, then give them ownership of the entire seeker electronics stack.

Ukraine accelerated the timeline. The company spun out of Defense Tech for Ukraine, a volunteer non-profit that designed hardware for frontline units. Eight weeks of field iteration in 2026 produced a 45-unit sale and a pipeline the founders value at $500 million. That loop — prototype, fly, break, fix, fly again — mirrors the Indiegogo crunch but with a different customer: a brigade commander who needs the seeker to work in fog, smoke, and electronic warfare, not a productivity enthusiast who wants a gentler nudge.

When Meta spends billions normalizing smart glasses and the Pentagon redirects billions from legacy programs toward drones and counter-drone systems, the same supply chain (PCB fabs, component distributors, RF test gear) serves both markets. Engineers who learned to route 60 GHz antenna traces on a four-layer flex board for a wearable can route 24 GHz seeker arrays on a rigid-flex stack for an interceptor. The form factor shrinks from 80 mm diameter to temple width; the physics stays identical. Vernius is betting that the talent pipeline now flows from consumer hardtech into defense hardtech because the problems have converged. The seeker head is just the next product spec.

The Frontline Testbed

Ukraine has absorbed more than 57,000 Shahed-type strikes since February 2022. That volume forced a domestic industry to solve the cost-imposition problem that Western air defense still treats as theoretical. By January 2026 Ukrainian forces downed a record 1,704 Shaheds in a single month; 70 percent of those kills came from interceptor drones, not guns or missiles. The STING interceptor, built by Wild Hornets, accounts for the largest share. Wild Hornets reports nearly 4,000 drones and other targets destroyed since May 2025, including the first confirmed kill of a Russian Geran-3 jet-powered variant and a Shahed carrying an air-to-air missile. No other system in any military has that record.

The STING platform is a compact quadcopter priced around $2,100 per unit. Production runs over 10,000 units a month as of March 2026. Operators cite an 80–90 percent hit rate depending on the unit, with best daily results reaching 100 percent — one launch, one destruction. The airframe cruises at 140–170 km/h, tops out at 280 km/h, and operates from sea level to 5,000 meters with a 7,000-meter ceiling. Maximum range is 37 km; tactical radius with return is 18.5 km. Endurance stretches to 15 minutes at cruise. Payload capacity sits at 500 grams. Training takes six days for new pilots, two days or less for qualified FPV operators. In April 2026 a STING launched from an unmanned surface vessel, the Magura USV configured as a floating battery carrying 18 interceptors, recorded the world's first sea-based drone-on-drone intercept. By September the STING S variant had destroyed a high-speed Geran-5 cruise missile and around a dozen Geran-4 jet-powered attack drones.

Vernius chose this environment for the Archimedes 1A's first live integration. Vernius reported it has secured $500 million in letters of intent for 2027 from defense manufacturers and military units across the United States, European Union and Ukraine. Eight weeks of radar-seeker development and field testing in Ukraine produced the 45-unit military sale and $500 million in letters of intent, according to Vernius. The seeker head mounts on the STING airframe, adding active 24 GHz radar that detects Shahed-class targets beyond visual range in fog, smoke, rain and darkness. The architecture accepts bearing, range and altitude data from a cueing network, provides active-radar midcourse guidance, and supports autonomous vision-based terminal guidance. The same hardware can feed target telemetry to a human operator for assisted rather than fully autonomous employment. That flexibility matters because Ukraine's interceptor doctrine still relies on pilot-in-the-loop engagement; Wild Hornets' own Alex Roslin told CBS News that an experienced pilot can learn their drones in three days, but building teams and an integrated network of air defense takes far longer.

The Ukrainian testbed validates three things no range can. First, the seeker survives the vibration, thermal cycling and electromagnetic noise of a $2,100 airframe launched from a van in an open field; the typical STING deployment uses modified civilian vans spaced three to five kilometers apart, 12 to 15 people per platoon. Second, the cueing network works: Ukraine's layered detection mesh of radars, acoustic sensors and human observers feeds the interceptor before it leaves the rail. Third, the cost target holds. A Patriot PAC-3 MSE costs over $13.5 million; Lockheed Martin produced 600 in all of 2025. Gulf states burned through more than 800 in three days during March 2026 Iranian strikes. The STING's $2,100 unit cost changes the arithmetic. Vernius bets the Archimedes 1A can keep the seeker subsystem cheap enough that the complete interceptor stays in that band.

Ukraine's export ban, in place since the 2022 invasion, means every STING built goes to Ukrainian forces. Wild Hornets cannot negotiate independently. Three Ukrainian expert teams (military personnel, engineers, drone specialists) have deployed to Qatar, the UAE and Saudi Arabia; a fourth supports U.S. forces in Jordan. Eleven nations have formally requested assistance. The constraint is not factory output. It is the legal framework and, underneath that, the political will in Washington to close a deal Kyiv has offered for eight months. Vernius, spun out of that nonprofit, works directly with operators and airframe partners from prototype through live-fire testing. That loop — build, fly, break, fix, fly again — is the only way to certify a seeker for mass production. The Gulf has weeks to field a capability Ukraine refined over two years. Vernius is proving the hardware can move at that speed.

Can They Build Fast Enough?

The founders put it bluntly on their Y Combinator page: they are building the seeker for mass production. That figure, call it 120,000 to 300,000 seekers annually, is not a typo. It is the production target implied by the cost arithmetic that has haunted air defense since Shahed drones began arriving in Ukraine at apiece while the missiles sent to kill them cost two orders of magnitude more. Defense Secretary Pete Hegseth framed the same problem in December: "We cannot afford to shoot down cheap drones with $2 million missiles. And we ourselves must be able to field large quantities of capable attack drones." His public goal: 300,000 drones "quickly and inexpensively," with "hundreds of thousands of them by 2027."

Vernius's that amount in letters of intent for 2027, from U.S., EU, and Ukrainian defense manufacturers and elite units, suggests the demand signal is real. But the gap between a letter of intent and a shipping pallet is where radar-seeker programs have historically died. The seeker head is often the costliest component in an autonomous interceptor, and Vernius's own materials acknowledge that its engineering pipeline is designed to move from requirements to production-oriented hardware faster than traditional cycles. That speed requirement collides directly with the PCB supply chain.

Echodyne, a radar startup backed by Bill Gates, offers the nearest public benchmark. In July 2026 it opened an 86,350-square-foot facility in Washington state, a $40 million investment targeting more than 30,000 radar units annually. Its Metamaterials Electronically Scanned Array (MESA) architecture was chosen specifically because it eliminates thousands of phase shifters, reducing size, weight, power, complexity, and cost. Even with that simplification, Echodyne's CEO Eben Frankenberg said the only way to defend against mass is with mass, and that requires "not just high-performance economical radars, but the ability to manufacture them at scale." Vernius is aiming for an order of magnitude more units on a form factor one-tenth the diameter.

The Archimedes 1A is an 80 mm, 24 GHz active radar seeker with a fixed beam. That packaging forces RF PCB design into a cylindrical volume where trace geometry, material selection, and thermal management interact in ways that standard FR-4 boards do not tolerate. Modus Advanced, a contract manufacturer that supports radar-seeker programs across six disciplines (precision machining, RF shielding, form-in-place gasket dispensing, thermal management, coatings, and soft-goods converting), notes that vertical integration under one roof cuts lead times and eliminates hand-off quality risks. Their standard CNC tolerance is ±0.25 mm; FIP gasket bead tolerance is ±0.15 mm. Vernius will need that class of control across thousands of units per month.

Then there is the compliance layer. As of November 2025, CMMC 2.0 Level 2 certification is a binding award criterion on new DoD solicitations. Modus holds that certification; any domestic PCB house Vernius uses must as well. The defense electronics supply chain already contends with foreign concentration, component obsolescence, counterfeit infiltration, and lead-time volatility that compound across program lifecycles. Vernius's job posting lists "shipped radar or phased array PCBs" and "published research, or shipped PCBs at scale" as bonus qualifications, a signal that the talent pool who can actually execute this volume is thin.

The company's pedigree helps. The founders spun the company out of that nonprofit, which designed hardware now used by Ukrainian units. That lineage means they have already moved boards from prototype to frontline use under wartime lead times. But wartime improvisation and peacetime rate production are different regimes. The LOIs are for 2027. The factory floor, whether owned or contracted, has to be ready before the first purchase order lands.

Incumbents Protect Margins; Startups Attack Cost

The counter-drone radar market is splitting along a familiar fault line: incumbents protecting high-margin programs while startups attack the cost curve. The numbers frame the fight. Fortune Business Insights' data shows the global counter-drone radar market at in 2025, projecting by 2034 at a 13.5% CAGR. Market.us's figures put the broader counter-drone market at in 2024, rocketing to by 2034, a 27.9% CAGR. North America owns roughly 48% of the radar segment. Hardware dominates at 69% of spend. Ground-based systems take 77%. Anti-drone radar itself captures 56% of the technology mix. Destructive mitigation (kinetic intercept) commands 82%. The military end-user sits at 62%. This is not a niche. It is a procurement wave.

Incumbents occupy the high end. BAE Systems signed a seven-year framework agreement with the Department of War to quadruple production of the infrared seeker for the THAAD interceptor, a program built for ballistic missile defense, not swarms of $20,000 Shaheds. Boeing is pushing its Universal Low-Cost Seeker (ULCS) architecture, explicitly designed to "dramatically reduce per-unit costs and enable precision munition employment at scale across programs." The company's hiring data on Zero G Talent shows 21 roles added in the past week alone (avionics, mission analysis, product security). That is a production engine spinning up, not a skunkworks. Lockheed Martin, Raytheon, Thales, Saab, IAI, and Airbus round out the tier-one list tracked by Market.us. Their seekers are exquisite, ITAR-controlled, and priced for platforms that cost millions per shot.

Startups are betting the economics invert when the threat is a thousand drones a week. Vernius Systems unveiled the Archimedes 1A, an 80mm, 24 GHz fixed-beam radar seeker, and demonstrated it on Wild Hornets' STING interceptor in Ukraine. The company says seeker heads can be the costliest component in an autonomous interceptor, and its pipeline uses an accelerated timeline. The target: tens of thousands of units a month.

They are not alone. Arbe Robotics entered the market in September 2026 with the Alerion C-UAS Radar, a compact 4D imaging radar now under evaluation by integrators worldwide. Saab developed the Loke mobile counter-drone system in 84 days, announced March 2025. DroneShield upgraded its systems at the Avalon Airshow the same month. Blighter Surveillance launched BlighterNexus, an AI-powered radar integration hub, in February 2025. The pattern: speed to field, software-defined architecture, and a price point that assumes attrition.

The U.S. Army alone requested over $500 million for counter-UAS in FY2025. The DoD is moving billions from legacy lines to drones and counter-drone. Ukraine produced over 2 million drones domestically in 2024 and claims capacity for 4 million annually, a production culture that rewards iterate-and-fire over perfect-and-procure.

Vernius sits at the intersection: a seeker designed for the interceptor, not the radar truck. Its 80mm form factor fits the kinetic layer that incumbents have largely ignored because the volume math never worked — until Shahed salvos made it work. The question is not whether startups can build cheaper seekers. It is whether they can qualify them fast enough to matter while Boeing and BAE defend the programs that pay the overhead.

Comparable Dollar Figures

The Factory Floor Waits for 2027

The factory floor that has to replicate that moment 300,000 times a year does not exist yet. Vernius has letters of intent, a Y Combinator batch, and a PCB engineer hiring pipeline filtered for people who have already shipped boards in anger. They have the Ukrainian testbed, the cueing network, the airframe partner producing 10,000 units a month. They have the physics (24 GHz, fixed beam, commercial fab) and the talent pipeline from consumer hardtech. What they need now is the same thing every defense startup needs: the first purchase order to land before the supply chain breaks, the compliance audit to clear, the thermal cycling to pass on the ten-thousandth board the way it did on the tenth.

The smart-glasses pivot was the proof. The seeker head is the product. The winter fog over Kyiv is the customer.

Working in frontier tech? Zero G Talent tracks the openings: see every open Boeing role, browse frontier tech jobs, the companies hiring, and the people building the field.