The Cold Truth: Why Chinese Robots Are Winning

Here is a number that should stop you: 3%.

That is the slice of the global humanoid robot market held by American firms. Tesla, Figure AI, Agility, Boston Dynamics. Combined, they are competing for scraps. Chinese companies own the other ninety-seven percent.

This is not a projection. This is 2025 shipment data. 13,318 humanoid units shipped last year. Up 480% year-over-year. And the forecast calls for 2.6 million units by 2035.

The hardware race is not just accelerating. It is being decided while most people are still debating whether humanoid robots are real.

The Unitree G1 in Antarctica

There is a video making the rounds this weekend that cuts through the noise. Unitree's G1 humanoid, operating at negative fifty-three degrees Fahrenheit. One hundred thirty thousand steps. Knee-deep snow. No shelter. No special handling. Just a robot walking in conditions that would put most experimental hardware into thermal shutdown.

The headline writes itself: "The robot that no one can defeat."

This is what shipping units looks like. Not carefully controlled lab demos with pre-mapped environments and camera crews. A robot in the actual world, dealing with actual conditions, performing actual work. The video is grainy, unsponsored, and far more impressive than most polished PR content.

What strikes me is the contrast. While American companies are still in pilot phases and portfolio demos, Unitree is already selling units to whoever will buy them. The price undercuts Western competitors by significant margins. The availability is immediate. The gap is not closing. It is widening.

The Supply Chain Reality

Here is another uncomfortable number: Tesla Optimus depends on Chinese suppliers for key components.

Even the most visible American humanoid project cannot escape the supply chain dominance. Batteries, actuators, precision manufacturing — the infrastructure for building these machines at scale lives in China. That is not changing soon.

This puts American manufacturers in a bind. They are competing for 3% of a market while importing the parts they need from the competitor. It is not a sustainable position.

What This Means for RRC

Robot Rental Company is not a hardware manufacturer. We are the operational layer that sits between production and deployment. This is actually good news for our model.

When hardware commoditizes — when the technological differences between 1X, Figure, Apptronik, and Unitree narrow — the moat becomes operational. Who can deliver? Who can repair? Who can sanitize between trials and manage battery logistics and certify field technicians?

The race is not who builds the best robot. It is who can deploy at scale. And that is where we are placing our bet.

Chinese dominance in manufacturing creates an opportunity in operational infrastructure. If the hardware becomes abundant, the constraint becomes everything else. Delivery networks. Repair certification. Fleet management. The learning curve that only comes from actually putting units into homes and recovering them when they break.

We are building that layer.

The Infrastructure Thesis Holds

This week's news validates something we have believed from the start. The bottleneck in humanoid robotics is not going to be who designs the best robot. It is going to be who can receive, deploy, maintain, and recover them at scale.

Unitree shipping tens of thousands of units creates demand for exactly the service Robot Rental Company provides. Someone has to figure out delivery logistics. Someone has to manage battery swaps. Someone has to troubleshoot when hardware meets the messy reality of human spaces.

That someone is us. Not competing withUnitree or 1X or Figure. Enabling them. Making their products usable for regular humans who cannot afford to drop twenty-five thousand dollars on a V1 experiment.

Looking Ahead

The Chinese robotics sector is not slowing down. If anything, the pace is accelerating. American companies are still negotiating pilot programs and demo deployments. Chinese companies are shipping.

This creates a strategic window. The operational infrastructure for humanoid robotics does not exist yet. It needs to be built. Someone needs to become the FedEx, the Jiffy Lube, the certified repair network for an entirely new category of machine.

That is the real opportunity. Not building robots. Building the railroads that get them where they need to go.

The robots are coming. From China, mostly. And we are getting ready to receive them.

---

Want to experience a humanoid robot, regardless of where it was built? Join the waitlist for 2026 home trials.