When Xpeng unveiled its next-generation Iron humanoid at its 2025 AI Day, its movements were so lifelike that over half of online comments questioned whether a real person was hidden inside. The CEO had to return to the stage 24 hours later and cut the robot open with scissors to settle the debate.
That won't always be an option. As robots move into factories, malls, and eventually homes, the question of how you tell them apart from people — and who's responsible when something goes wrong — stops being philosophical and starts being a paperwork problem. China has already answered it: every humanoid robot manufactured in the country now gets a 29-digit ID before it leaves the factory.
The 29-Digit "Birth Certificate"
China has begun implementing a mandatory digital ID system. The ID isn't just a sticker slapped on for recall purposes. This "birth certificate" is structured into four components to ensure global and local traceability:
- National Code (2 digits): Designed to track cross-border shipments and international sales.
- Manufacturing Firm (4 digits): Identifies the specific company responsible for the robot's creation.
- Product Model (6 digits): Categorizes the specific version; currently, over 200 models are already represented in the registry.
- Individual Serial Code (17 digits): Provides a unique identifier for every single unit produced.
By assigning these IDs, the state is formalizing a framework where robots are treated as individual actors within a legal landscape, much like aviation authorities tracking individual aircraft tail numbers or financial regulators recognizing "legal persons."
Surveillance from the Factory to the Junkyard
Central to this oversight is the Humanoid Full Lifecycle Management Service Platform, run by a standardisation committee under China's Ministry of Industry and Information Technology, out of the Hubei Humanoid Robot Innovation Center in Wuhan.
The tracking begins at the point of production and persists through the robot's entire operational life, ending only at a certified recycling facility. This platform maintains a continuous data relationship with the machine, monitoring maintenance logs, application scenarios, intelligence levels, and real-time telemetry for things like joint wear, battery status, and operational accuracy.
For an industry heading into factories, shops, and eventually homes, this data trail is what makes liability calls possible when something goes wrong.
Finding the "Gaps" in the Human Face
As robots move into public spaces, there's a practical problem: how do you give a robot a face that couldn't be mistaken for a real person? Researchers are developing Biometric Identity Provisioning (BIP) to answer it. The method maps the mathematical space that human faces occupy, then finds the gaps where no real face exists. A robot gets assigned one of those gaps as its biometric identity, so it can never accidentally match a real person in a facial recognition system.
The tool doing this is called GapGen. The concept is straightforward: find the coordinates on the map of human faces that nobody lives at, and put the robots there.
"Humanoids are something completely new for robotics, especially when it comes to standards. The human-centric design raises our expectation levels." — Aaron Prather, Director of Robotics & Autonomous Systems Programs, ASTM International
The Home Robot Dilemma: Appliance or Citizen?
The integration of these entities into daily life is creating ownership questions nobody has a clean answer to yet. Take 1X's Neo — a home robot already shipping to customers. Is Neo a simple product where regulation ends at the point of sale? Or is it a regulated entity subject to lifetime monitoring? The state's interest isn't just safety; it is the ability to perform rapid fault diagnosis and determine liability when something goes wrong.
Recently, a humanoid robot developed by a Huawei subsidiary broke records at the E-Town Humanoid Robot Half-Marathon, finishing faster than any human has ever run the same course. When a machine can outrun a person, the liability question when it malfunctions stops being theoretical. What happens when it doesn't just fail — but causes harm?
China's Global "Standard-Setting" Sprint
China isn't just winning the hardware race. In 2025, approximately 85% of the 17,000 humanoid robots shipped globally were manufactured in China. With over 140 Chinese firms developing humanoids and 28,000 robots already assigned digital IDs, China is building the de facto global standard by sheer volume.
The problem is that the rest of the world hasn't agreed to that. Western governments are unlikely to embrace an identification scheme that routes continuous operational data through Chinese state infrastructure. And without an international alternative, as industry analysts at Six Degrees of Robotics note, there are currently "no unified norms for traceability, safety supervision, or data circulation." A robot built in Shenzhen, tracked on a Chinese platform, operating in Berlin or Detroit is a jurisdictional puzzle nobody has formally solved yet.
Conclusion: Faster Than the Speed of Law
The ID system, the lifecycle platform, the biometric gap-filling — together, they represent something more than robot regulation. They are a governance architecture: a system for numbering, tracking, and monitoring physical agents moving through public and private space.
That architecture works. It also has a history. The same logic — assign an ID, track the unit, log its movements and behaviours — has been applied to human populations before, with consequences that weren't always about safety.
China is building this for robots because robots are new and the liability questions are real. But the infrastructure doesn't care what it's tracking. That's worth sitting with.