For decades, American industry quietly ceded the art of making things to China, trading manufacturing depth for cost efficiency. Now, as the US government moves to sever those ties in the name of national security, companies and universities are discovering that supply chains are not merely logistical arrangements — they are repositories of accumulated human knowledge and capability. The reckoning unfolding across robotics labs, university classrooms, and factory floors is less a trade dispute than a confrontation with the long consequences of forgetting how to build.
US push to ditch Chinese tech risks higher costs, slower innovation
We just don't know how to make them in the US.
So the core problem is that American companies built their entire supply chains around Chinese manufacturing. Now they're being forced to rebuild, and they can't.
Exactly. And it's not just about cost. When Bhageria tried to find someone to make plastic grabbers in the US, machine shops told him they literally don't have the capability. The infrastructure doesn't exist anymore.
But wait—how much of this is actually a capability problem versus a volume problem? If you're making a million units a year in China, the per-unit cost is tiny. In the US, you might be making ten thousand units. Of course it costs more.
That's fair. Murray at Kopin said the cost is 25% higher, but he also said China produces more micro displays in a week than the US produces in a year. So it's both—capability and scale.
And the universities are getting hit too. Mitra said replacing Chinese humanoids could cost ten times more. That's not a small difference.
Ten times is a big claim. Is that his estimate, or is that what he's actually seeing in quotes from vendors? The article doesn't say.
It says "could cost," so it sounds like an estimate. But the point stands—if universities can't afford the tools, they can't train the next generation of roboticists.
So we're looking at higher costs for companies, slower innovation in research, and consumers paying more for everyday things like robot vacuums.
And we don't actually know how long the transition takes. Kenneally says long-term the bans might work, but "long-term" could be five years or fifteen years. The article doesn't pin that down.
True. But Farid's warning is pretty stark—if alternatives don't get created fast enough, it could derail the whole goal of building a domestic robotics industry.
Because companies will just fall behind China while they're waiting for American suppliers to catch up.
Exactly. The policy assumes American manufacturing can scale up quickly. But there's no evidence in this story that it can.
Il Polso
- American robotics and tech firms are scrambling to replace Chinese components they depend on, only to find that domestic alternatives are often more expensive, limited, or simply do not exist yet.
- The gap is staggering: China produces in a single week what the entire US manages in a year for certain critical components, exposing how deeply decades of outsourcing have hollowed out domestic manufacturing capacity.
- Restrictions are expanding faster than solutions — rare earth metals, micro displays, plastic grabbers, and consumer robots are all caught in a widening net, with bans arriving before American suppliers can fill the void.
- Universities and researchers face a quiet crisis too, as Chinese-made robots and drones used in classrooms could cost ten times more to replace with American equivalents — if replacements exist at all.
- Industry leaders are split between cautious optimism that the pressure will eventually stimulate domestic growth and urgent fear that the transition period will cost America its competitive edge in innovation.
For decades, American industry quietly ceded the art of making things to China, trading manufacturing depth for cost efficiency. Now, as the US government moves to sever those ties in the name of national security, companies and universities are discovering that supply chains are not merely logistical arrangements — they are repositories of accumulated human knowledge and capability. The reckoning unfolding across robotics labs, university classrooms, and factory floors is less a trade dispute than a confrontation with the long consequences of forgetting how to build.
Rajat Bhageria built Chef Robotics on a simple premise: source the best components at the best price. Like most American robotics firms, that meant relying heavily on Chinese manufacturing. But the question investors and customers now ask him has changed — and so has the landscape he operates in.
The Trump administration has spent the past year expanding restrictions on Chinese technology, adding power inverters, advanced robotics, and consumer devices to a growing list of banned or heavily tariffed goods. The stated rationale is national security and the revival of domestic manufacturing. The practical reality is that American companies are being asked to solve a problem they have never had to confront.
When Bhageria tried to source plastic grabbers for his robot arms domestically, American machine shops told him plainly: the capability does not exist here. Michael Murray of Kopin Corporation spent over two years relocating micro display production from China to the US, absorbing a 25% cost increase — and still operates in a country that produces in a year what China makes in a week. Gavin Kenneally of Ghost Robotics has moved motor production to South Korea and magnets to Europe, but the raw rare earth materials still come from China, placing him on a collision course with a January ban.
MIT's Ben Armstrong framed the deeper problem: decades of outsourcing left America without the knowledge, not just the factories, to rebuild quickly. The learning curve is real, and during it, prices will be high and innovation will slow.
At the University of Illinois, engineering professor Sayan Mitra watches this unfold in his own classrooms, where Chinese-made robots and drones are central to coursework. American replacements could cost ten times more — or may not exist at all. He compared the situation to performing surgery on a patient already standing: done carelessly, it risks harming the very system it means to heal.
Whether the pressure ultimately stimulates a genuine domestic renaissance or simply widens the gap between American ambition and Chinese capacity remains the defining question — one that companies, universities, and consumers will be living with for years to come.
Rajat Bhageria founded Chef Robotics seven years ago in San Francisco with a straightforward business model: build robot arms that food manufacturers could use to automate meal assembly, sourcing components wherever they were cheapest and best. Like nearly every robotics firm in America, his company became dependent on Chinese parts. China's factories could produce components at volumes and prices no competitor could match. But that calculus has shifted. Investors and customers now ask him a different question than they did before: Why are you still using foreign-made parts?
The Trump administration has spent the past year systematically restricting what American companies can buy from China. The list keeps growing. In July, the Federal Communications Commission added power inverters and advanced robotics—humanoids that can run, jump, dance and fight—to the banned technologies. The government has imposed a 100% tariff on Chinese electric vehicles, similar tariffs on drones, and forbidden cars that use Chinese software. Even Ford, one of America's largest manufacturers, drew criticism from the administration for using Chinese electric car batteries made by CATL. The stated rationale is national security: protecting sensitive data and incentivizing domestic manufacturing. But the practical effect is forcing American companies to solve a problem they have never had to solve before.
Bhageria's robot arms are not yet affected by the bans, but he is moving now anyway, trying to get ahead of what he expects will be an expanding list of restrictions. He discovered quickly that the plastic grabbers attached to the end of his arms—currently made in China—cannot be produced cheaply in the United States. When he approached American machine shops, they told him plainly: we cannot do this. The capability does not exist here. "It's actually been a much harder process than we had initially hoped," he said.
The challenge Bhageria faces is not unique. Michael Murray runs Kopin Corporation, which makes optical components for military drones. His company spent more than two years relocating manufacturing for micro display screens from China to the United States. The cost of making those components domestically runs about 25% higher than it did in China, though the Department of Defense has offset some of the expense. Murray noted a stark disparity: China can produce more micro displays in one week than the United States can produce in a year. Gavin Kenneally, whose company Ghost Robotics makes industrial and military robot dogs, is searching for new sources of rare earth metals. China controls nearly the entire global supply. He has already moved motor production to South Korea and sources neodymium magnets from Europe, but the raw material still comes from China, which means it will be banned starting in January.
Ben Armstrong, executive director of MIT's Industrial Performance Center, explained the underlying problem. "A lot of our digital lives are built on the infrastructure of foreign-made components and goods assembled abroad, and that's not going to change anytime soon," he said. "We just don't know how to make them in the US. So, there will be a learning curve, and during that learning curve process, the price will be high." Decades of outsourcing have left American manufacturers without the capacity to meet demand. The result is not just higher costs for companies; it is slower innovation and reduced access to cutting-edge tools.
At the University of Illinois Urbana-Champaign, engineering professor Sayan Mitra said several of his courses rely on robotic arms, drones, humanoids and vehicles from China. Replacing Chinese humanoids with American equivalents could cost ten times more. Other drone options may not exist at all. "It's like we are trying to do surgery on a patient that is already on their feet," Mitra said. "If it is done in a very ad hoc, unpredictable way, it's going to slow down everything." The FCC ban on mobile robots includes new versions of automated devices like robot vacuums, pool cleaners and lawn mowers—the vast majority of which are made in China. Jitesh Ubrani, director of consumer devices research at IDC, said moving that manufacturing to the United States would be extremely difficult and very expensive. American consumers will end up paying significantly more.
Saman Farid, founder of the industry coalition Robots for America, worries that restrictions are expanding faster than alternatives can be created. "In the time that it takes an American company to build their first prototype, a Chinese company may already be on version three," he said. Gavin Kenneally acknowledged that the supply chain problems are extremely difficult to solve in the short term, but he believes the bans may eventually have their intended effect. "In the long term, I think these bans have had the stimulative effect that the government had designed them to have." The question is whether American companies and universities can survive the transition without falling too far behind.
Citazioni salienti
It's actually been a much harder process than we had initially hoped. We will even talk to some really good machine shops in the US and they're like, 'We can't do this.'— Rajat Bhageria, founder of Chef Robotics
In the time that it takes an American company to build their first prototype, a Chinese company may already be on version three.— Saman Farid, founder of Robots for America