Chinese Robots Lack a 'Brain' and Data: Embodied AI's Biggest Challenge at WAIC 2026

Chinese robot makers at the World Artificial Intelligence Conference (WAIC), which wrapped up in Shanghai on July 20, 2026, named a shortage of data and the lack of a strong 'brain' — a foundation model for interacting with the physical world — as the main obstacles to embodied AI.

What Chinese robots are missing

Chinese robots are short on two things at once: high-quality data about the physical world and a strong foundation model that ties perception to action. Industry representatives said as much at WAIC, and, according to the South China Morning Post, this double deficit is what keeps embodied AI from moving from demos to real products.

Unlike language models, which learn from text scraped from the internet, robots need data on movement, grasping objects and reacting to an unpredictable environment. That data has to be collected by hand, which is expensive and slow. Without a large volume of it, a robot's 'brain' cannot learn to act confidently outside the lab.

• Event — the WAIC conference, which ended on July 20, 2026 in Shanghai
• Who is speaking — Wang Xiaogang, co-founder of SenseTime and chairman of its robotics division
• Barrier 1 — a lack of data on interaction with the physical world
• Barrier 2 — the absence of a strong 'brain', a unified foundation model
• The task — to link hardware, data, models and scenarios into a closed loop

What is a 'closed loop'?

WAIC participants called the most critical challenge for the embodied AI industry building a closed, iterative system in which hardware, data, models and real-world scenarios reinforce one another. The idea: a robot collects data while it works, that data improves the model, the improved model makes the robot more effective, and the cycle repeats.

"To link hardware, data, models and real-world scenarios into a closed iterative system," said

Wang Xiaogang, co-founder of SenseTime and chairman of its robotics division, describing the industry's main task.

Right now that loop is broken. Most Chinese companies have hardware platforms, but they have neither a sufficient stream of data from real-world operation nor a model capable of fully digesting that data. Until the links are connected, each prototype remains a one-off demonstration rather than a scalable product.

Why is data the bottleneck?

Data is the main bottleneck for robots because it cannot simply be downloaded from the web the way text or images can. Every action in the physical world — a step, a turn, grasping a cup — has to be recorded through the sensors of a real device or through costly simulation. The volumes involved are not comparable to the trillions of tokens that language models train on.

That is why WAIC participants stressed the pairing of 'data + brain': a powerful model without data will not learn to act, and a huge pile of data without a unified foundation model will not turn into meaningful behavior. By the estimates of conference insiders, the industry is still looking for a way to put data collection on a production line so that it pays back the cost of the hardware.

What this means

Chinese robotics is bottlenecked not by hardware but by intelligence and data. The country is already strong on the hardware side, but without a 'brain' and a stream of real-world data, robots do not leave demo mode. Closing the hardware–data–model–scenario loop is becoming embodied AI's main race for the next several years.