On a patch of dirt in the vast wilderness in Alaska, a long-range drone roared like a lawn mower as it shot into the sky. It scanned the ground for a target it had been programmed to recognize, and then dived, attempting to destroy it by crashing into it. But it missed, landing about 80 feet away.
On another attempt, a drone nose-dived at launch. On a subsequent try, a drone crashed into a mountain.
These drones weren’t flown by amateur hobbyists. They were launched by drone manufacturers paid by a special unit of the Department of Defense as part of an urgent effort to update U.S. capabilities. For four days last month, they tested prototypes of one-way drones by trying to crash them into programmed targets, while soldiers tried to stop the drones with special electronic equipment.
The exercise aimed to help U.S. defense contractors and soldiers get better at drone warfare. But it illustrated some of the ways in which the U.S. military could be unprepared for such a conflict. The nation lags behind Russia and China in manufacturing drones, training soldiers to use them and defending against them, according to interviews with more than a dozen U.S. military officials and drone industry experts.
“We all know the same thing. We aren’t giving the American war fighter what they need to survive warfare today,” said Trent Emeneker, project manager of the Autonomy Portfolio at the military’s Defense Innovation Unit, which organized the exercise in Alaska and paid for the development of the drone prototypes that flew there. “If we had to go to war tomorrow, do we have what we need? No. What we are trying to do is fix that.”
Defense Secretary Pete Hegseth has acknowledged that the country has fallen behind, and he announced a series of new policies and investments in drones that he vowed would close the gap. In a video released on Thursday, he cited outdated rules and procurement processes as making it too difficult for commanding officers to buy drones and train their soldiers to use them.
“While our adversaries have produced millions of cheap drones,” he said, “we were mired in bureaucratic red tape.”
The video came on the heels of an executive order signed by President Trump last month called “Unleashing American Drone Dominance,” which directs federal agencies to fast-track approvals for American drone manufacturers and protect the U.S. drone supply chain from “undue foreign influence.”
But it will take time and money to grow a domestic industry capable of producing enough drones to meet the needs of the U.S. military. Although the United States has excelled in developing large, complex unmanned aircraft like the Predator and Reaper drones, which cost tens of millions of dollars apiece, today’s conflicts have been dominated by swarms of smaller, inexpensive drones that are largely produced with components from China.
The Defense Department did not respond to requests for comment.
Drones have become a weapon of choice on modern battlefields. In the early days of the war in Ukraine, soldiers beat back the Russian invasion by adding deadly modifications to the Mavic, a drone sold to hobbyists by DJI, a Chinese company that is the world’s largest drone manufacturer. Versions of the Mavic cost between $300 and $5,000, according to online retailers.
DJI, of Shenzhen, China, accounts for about 70 percent of all commercial drones sold globally for hobby and industrial use, such as aerial photography, package delivery and weather research. The privately held company sells its equipment to customers in the United States — there’s even an authorized store on Fifth Avenue in Manhattan — but U.S. law bars the military from buying Chinese drones. The company declined to share market data, but industry experts estimate that DJI’s output far exceeds that of any other drone manufacturer.
“No one even comes close,” said Bobby Sakaki, chief executive of UAS NEXUS, a drone industry consultant. “DJI can make millions of drones per year. That is a hundred times more than anybody in the United States can make.”
Although DJI is not a military company and said it cuts off customers who use its drones for armed conflict, its near-total dominance of the market for drones and drone components has caused alarm in Washington, where some lawmakers want to ban its products so that a domestic industry can flourish.
But it will take time and money to grow a domestic industry capable of producing enough drones to meet the needs of the U.S. military. Enter Silicon Valley investors who have been pouring money into American drone companies, anticipating that the Defense Department will place a large order for American-made drones. Peter Thiel’s Founders Fund has invested more than $1 billion in Anduril Industries, an American defense technology company that specializes in advanced autonomous systems. Mr. Trump’s son Donald Trump Jr. joined the board of Unusual Machines, another U.S. drone maker, last year.
About 500 companies manufacture drones in the United States, producing fewer than 100,000 a year, according to Ryan Carver, communications manager for the Association for Uncrewed Vehicle Systems International, a nonprofit organization of industry professionals. But many are start-ups without a track record of production or sales. Founders jockey for the chance to show off their wares to military units that are beginning to work with drones. The changes announced by Mr. Hegseth on Thursday, which make it easier for commanders to buy drones, will intensify that competition.
“Everyone wants to win the Army’s big drone contract, get their billion-dollar check and go retire on an island somewhere,” said Nathan Ecelbarger, chairman of the U.S. National Drone Association, which promotes the rapid advancement of drone and counter-drone technologies.
But the exercise in Alaska showed how hard it can be to develop homegrown drone capabilities.
The first two days of testing were full of setbacks. Two companies were testing prototypes of a long-range unmanned craft that could fly for hours, navigate without GPS or a human operator, and crash into a target that it had been trained to recognize. They were among four finalists — out of more than 100 applicants — to get the money from the Defense Innovation Unit to develop the systems. Two other companies were set to test their prototypes in Ukraine.
The craft made by Dragoon, a start-up in Tucson, Ariz., experienced engine trouble and then issues with navigation. It failed to hit a target. But on the final day, it recognized a target — an M113 armored personnel carrier — and swooped down to crash into it. The hit was considered a success, even though the target had not been the one intended.
“We have got a lot of work to do to make it operational, for sure,” said Jason Douglas, one of three co-founders of Dragoon. “But those were huge steps.”
AeroVironment fared worse. At first, its drone failed to launch. Then one crashed into a mountain after its navigation system was blocked, narrowly missing a group of soldiers who stood with their jamming equipment. Although one of its drones flew long distances, and successfully crashed into a target with the help of GPS, the prototype never hit a target once its GPS was blocked.
Paul Frommelt, a spokesman for the company, noted that the exercise was a chance to collect data on “an experimental variant of one of our products.”
While many small drones are controlled by human operators, the Defense Innovation Unit has been trying to develop semiautonomous systems that can be trained by artificial intelligence to recognize targets — enemy tanks, for instance — and attack them even if communication with the human operator is cut.
“Do we need a capability like this? Yes. I think that is very clear in the modern battlefield,” said Mr. Emeneker, who is a civilian contractor for the D.I.U. But he acknowledged that the project might not succeed. “Things haven’t gone as smoothly as I wanted. It’s clear that there are some still really hard technological challenges to overcome.”
The soldiers who participated in the exercise, most of them from the Electromagnetic Warfare Platoon of the Army’s 11th Airborne Division, experienced their own problems. On a mountaintop, they set up six tall electronic jammers, which looked like slender microphones attached to black tripods. They emitted radio signals that were supposed to overpower the signals sent by the drone operators. But those jammers — some of which were designed more than a decade ago to fight the war on terror — had hardly any effect. Neither did the backpacks containing newer drone-disarming equipment that some soldiers wore.
The team had a drone-buster, too — a huge gunlike device that looked like something from the movie “Ghostbusters.” But no one bothered to try it. “That thing never worked,” one man said.
Over time, the soldiers improved. By the fourth day of the exercise, they had figured out how to use their jamming equipment more effectively. A black suitcase-sized box called a Magpie worked particularly well, they said.
But Lt. Col. Scott Smith, director for the nonlethal effects section of the 11th Airborne Division, said the exercise highlighted how much more work Americans needed to do to prepare for a conflict involving drones.
“Their equipment just doesn’t have the desired effect against the latest technology,” he said.
Chris Bonzagni, a drone industry consultant with Contact Front Technologies who helped put on the Alaska test, said many of the American drones that were initially delivered to Ukraine failed on the battlefield because they were outdated or easily jammed by the Russians.
“In Ukraine, the companies delivering tech to the war fighters are with them all the time, observing firsthand what is working and what is not,” he said. Ukraine has also become a drone-making hub because its soldiers and engineers are forced to master drone technology to survive, something Americans have not experienced yet.
The event was held at the Yukon Training Area, a military site about an hour south of Fairbanks, because it was the only place where organizers could get permission to fly the drones while soldiers tried to try to jam them, Mr. Emeneker said. The electronic signals used to disrupt drones can wreak havoc with civilian aircraft, radios and cellphones, making it difficult to get clearance to conduct such exercises in populated areas.
Some U.S. drone companies do their testing and development in Ukraine. That may be why one drone stood out in Alaska: a small, short-range quadcopter created by Neros, a start-up in El Segundo, Calif., with an office in Ukraine, which was testing a radio.
That drone, called the Archer, managed to hover about 10 feet over the soldiers’ heads, despite their jamming equipment. Its radio toggled between multiple frequencies, switching every time soldiers tried to jam it. It carried a jar of strawberry jam, a joke from Soren Monroe-Anderson, the 22-year-old chief executive and co-founder who piloted it with what looked like a video game controller.
Neros, which is providing about 6,000 drones to Ukraine this year and produced the drone that appeared in Mr. Hegseth’s video, has been described by some U.S. military leaders as their best alternative to China’s DJI. The Archer sells for about $2,000 each, making it one of the most affordable models. But Neros produces only about 1,500 Archers per month in a factory where 15 workers assemble them by hand.
Mr. Monroe-Anderson, a former drone racer, said he was ramping up production and wanted to build a factory capable of making a million drones a year. He aspires to compete with DJI, but acknowledged the daunting odds.
“It is so much better than really any other company in the world,” he said of DJI. “That’s the reason why it is important to do what we are doing.”
The post Drones Are Key to Winning Wars Now. The U.S. Makes Hardly Any. appeared first on New York Times.
