WASHINGTON — The Pentagon’s acting head of the Defense Innovation Unit has given the United States’ effort to mass-produce cheap military drones a stark assessment: the country is “starting from basically zero.” The remark, reported in a Bloomberg feature published on Saturday, October 31, 2026, and summarised the same day, frames a race in which American factories are still learning to build at scale while rivals are already producing by the millions.
Travis Metz, the acting head of the unit charged with bringing commercial technology into the military, put the Drone Dominance Program’s goal at about 250,000 drones by the end of 2027, backed by roughly 1.1 billion dollars the Pentagon has committed to a four-stage competition launched by Defense Secretary Pete Hegseth in July 2025. That target, ambitious by American standards, is a fraction of what Russia is expected to build in a single year.
According to Ukrainian intelligence cited in the reporting, Russia is expected to build 7.3 million small expendable drones this year. China can surge to 500,000 a month, according to a Bloomberg Intelligence report from February, and Ukraine says it will make 10 million this year and plans to double that. The contrast is the point of the October 31 report, which tours factories in Utah meant to close the gap — and finds them, for now, producing in the hundreds and thousands, not the millions.
The report does not claim that American drone technology is inferior. It claims something more specific, and in some ways more difficult to fix: the United States has plenty of companies that can design a capable drone, but almost none that can build the same drone hundreds of thousands of times, on schedule, with parts that do not depend on a foreign supplier. That distinction — between inventing and manufacturing — runs through every number in the story.
What the Pentagon’s drone target actually is
The Drone Dominance Program is the Pentagon’s main vehicle for buying small, low-cost, one-way attack drones — the kind of aircraft that, in Ukraine, have become consumable battlefield equipment rather than precious platforms to be protected. The program is structured as a competition in stages, known as Gauntlets, in which companies compete for production orders.
The first phase, Gauntlet I, paid 11 companies 150 million dollars in March for 30,000 one-way attack drones, due over five months. By mid-September, only three of those companies had delivered every drone on the program’s public leaderboard, according to the reporting. Pentagon officials put the average price in that round at about 5,000 dollars per drone — cheap by the standards of fighter jets and submarines, but still a figure that only makes sense if production reaches volume.
One company, Neros Technologies of Torrance, California, picked up 800 drones that rivals failed to deliver on time, and also won an Army contract in June worth up to 500 million dollars for first-person-view drones, according to government contracting data cited by Bloomberg. A second round, Gauntlet II, covers 60,000 drones and produced 10 winners in mid-September, including Vector Defense of Bluffdale, Utah — the company the Bloomberg feature opens with.
Vector makes a few hundred small quadcopters a month. Its chief executive, Andy Yakulis, describes the product as ammunition: “a mortar round or a bullet or a hand grenade that just happens to fly.” The line is revealing, because it reframes the whole procurement problem. Ammunition is bought by the million and used without a second thought. A drone bought through a tournament, with winners paid months later, is not yet being treated that way by the institution buying it.
Readers should note that different reports have described the program’s overall ambition in different terms — some citing more than 200,000 or 300,000 drones over the life of the program. The October 31 report’s figure of about 250,000 by the end of 2027 is the one Metz is reported to have given for the current goal, and it is the figure used here. The variation reflects a program whose targets have shifted as it has moved through its stages, not a single fixed number set in stone.
Why the Pentagon cheap drone production gap is so wide
The gap is not only about factory floor space. It is about the supply chain underneath the factory, and that chain still runs, in large part, through China. China dominates the supply of batteries, brushless motors and propellers for small drones, and Pentagon rules bar or restrict Chinese parts in military systems. That is a deliberate security choice — a weapon that must work when China is the adversary should not depend on Chinese components — but it makes every compliant part more expensive and harder to source.
Teal Drones, in South Salt Lake, builds 1,000 Black Widow and 2,000 Fang first-person-view drones a month, and pays a premium for every compliant part. Chris Ericson, chief operating officer of Teal’s parent company, Red Cat Holdings, told Bloomberg that motors and cameras, along with radios and batteries, are the most expensive parts on a drone, and that “you can buy them cheap in China.” Teal pays the premium instead. Skydio chief executive Adam Bry says his company has effectively eliminated direct dependence on China, but warns that “anybody who says their dependence on China is purely zero is probably deluding themselves,” because suppliers sit several layers below the companies that assemble the aircraft.
The Pentagon’s answer is to pay for component plants rather than only for finished drones. Huntsville-based Performance Drone Works holds 820 million dollars in Defense Department loans, and aims to turn out 1 million critical components a year by the end of 2028 from a facility that today can build 10,000 complete drones annually. The Army has pivoted the same way: its plan last year was a network of factories inside its own ageing industrial base producing 10,000 drones a month with 3D printing, but the effort now focuses on parts, with a target of 50,000 brushless motors by the end of fiscal 2027.
That shift — from counting drones to counting motors, batteries and radios — is the most important change in the program’s logic. A drone company can promise thousands of aircraft, but if it cannot get compliant motors at volume, the promise cannot be kept. Industry executives have made the same point in trade-press interviews: America does not lack drone companies, it lacks depth underneath those companies, and predictable demand is what gives suppliers the confidence to invest in factories and automation.
The Army also opened an online drone marketplace in March, where units order drones and soldiers leave reviews, and the Pentagon expanded it to the other services in August. Colonel Danielle Medaglia, the Army’s uncrewed aircraft systems project manager, told Bloomberg that more than 850 million dollars has flowed through it. Yet the Army itself has bought only 450 small drones through that marketplace, from 21 vendors, a figure Bloomberg does not reconcile with the larger total. Separately, the Army took delivery of 8,500 small drones through Gauntlet and service programs in fiscal 2026. The marketplace, in other words, is still more promise than pipeline.
The factories, the money, and the midterm question
Private money has moved faster than Pentagon orders. More than 60 billion dollars went into over 220 American companies from the start of 2025 through September, and more than 80 of them announced new or expanded plants totalling 15.8 million square feet, according to the Association for Uncrewed Vehicle Systems International, as cited in the reporting. AeroVironment is targeting 400 Switchblade drones a month from a new Salt Lake City plant by next year, and Anduril opened its Ohio industrial park earlier this year.
The public money on the table is larger than anything the previous administration’s drone effort had. The Biden administration’s Replicator initiative promised “multiple thousands” of autonomous attack drones within 24 months of its August 2023 launch, and delivered several hundred, according to a person familiar with the effort who spoke to Bloomberg. The Trump administration kept parts of Replicator, signed an executive order on domestic drone production in June 2025, and put tariffs of up to 100 percent on imported drones and components in August.
The Defense Department’s proposed budget for the next fiscal year, still awaiting Congress, carries about 55 billion dollars for autonomous warfare development, roughly 17 billion of it for uncrewed systems, by Bloomberg Intelligence’s count. That funding is not yet law, and Bloomberg notes that a Democratic majority in either chamber after the midterms could put it at risk — a reminder that industrial policy on this scale depends on politics as much as engineering. The fight for Congress is therefore not a backdrop to the drone program; it is one of the variables that will decide whether its budget survives.
The Pentagon’s own recent history is also part of the story. The department has been shaken by personnel turmoil in the run-up to the midterms, including a separate matter in which it referred a fired top planner to the Justice Department over alleged classified leaks. That case is unrelated to drone manufacturing, but it illustrates the institutional strain under which the drone effort is being run — a department trying to build a new industrial base while its senior ranks are in flux.
Metz’s assessment — that the United States has plenty of entrepreneurial drone companies and none of them with scale — is therefore less a criticism of any one company than a description of the starting point. Neros told DroneXL in August that it was already building 1,200 drones a week and promised a million a year by 2028. Teal runs 3,000 a month. Vector says it can go to thousands. The Pentagon’s entire two-year target is 250,000, and the Army has bought 450 drones through the marketplace it built so that units could order drones the way they order boots. An Army official called the Pentagon’s drone demand “artificial” only last week, and makers in that reporting asked for a year of orders, not another round of competition.
What to watch next
The next real test, according to DroneXL’s reporting, is Gauntlet III’s 400-drone production deadline, reported in September — a test of whether anyone outside Neros and Teal can ship at volume. After Gauntlet I, when eight of 11 vendors had not delivered everything by mid-September, scepticism is warranted. But the test is also the point of the competition: to find out, before a war rather than during one, which companies can actually deliver.
The numbers to hold the program to, on the reporting’s own logic, are not only the drone counts. They are the component numbers: Performance Drone Works’ 1 million parts a year by the end of 2028, and the Army’s 50,000 brushless motors by the end of fiscal 2027. If those plants deliver, the drone counts can follow. If they stall, no amount of competition design will produce aircraft at scale.
Retired Marine aviator Trent Emeneker, who helped vet drone makers for the Pentagon from 2020 until March, put the wartime requirement bluntly: in a major combat scenario, “we need to make a million a month,” and the United States needs to do it yesterday. Jan Osburg, an adjunct senior engineer at the Rand Corporation and lead author of a paper this year on what the United States needs to do to realise the drone revolution, called the administration’s moves several big steps in the right direction and “both a sprint and a marathon.”
Both descriptions can be true. The sprint is the 2027 target and the Gauntlet deadlines. The marathon is the component plants, the supplier base, and the years it takes to build an industrial ecosystem that China spent decades assembling. The October 31 report’s contribution is to make that marathon visible — to show, factory by factory in Utah, how far the United States still has to run, and how much of the race will be decided not by the drones themselves, but by the motors, batteries and radios inside them.
Sources
USN reviewed reporting from multiple outlets for this story. Read more from DroneXL, The Flying Engineer and Advanced Manufacturing.