Newsletter N°90 - June 2026 

China is accelerating the development and deployment of humanoid robots and embodied artificial intelligence, with the ambition of moving the technology from laboratory demonstrations to practical applications across multiple industries.

 

In June 2026, China's Ministry of Industry and Information Technology (MIIT) and the State-owned Assets Supervision and Administration Commission (SASAC) launched a nationwide initiative aimed at accelerating the adoption of humanoid robots in real working environments. Under this program, local governments and state-owned enterprises must submit implementation plans by the end of June and report their progress by November 2026.

 

The government has set ambitious objectives: by the end of 2026, more than 10,000 humanoid robots should be deployed across China and over 100 high-value application scenarios should be established. The targeted sectors include manufacturing, warehousing, logistics, retail, healthcare, workplace safety, equipment inspection and maintenance, emergency response, and disaster relief. According to Chinese authorities, real-world deployments will generate valuable operational data that can be used to improve embodied AI systems, enhance robot performance, and accelerate large-scale commercialization.

 

To support this national strategy, several embodied intelligence training facilities are being developed. One notable example is the "Embodied Intelligence Training Center" operated by "China Xiongan Group Digital City Technology Co., Ltd." in Xiong’an New Area, Hebei Province. The center serves as a large-scale training ground where robots learn to perform tasks in realistic environments before being deployed in the field.

 

At the facility, data collection technicians remotely control robots using handheld controllers to perform various tasks. The resulting data is then used to train large AI models, enabling robots to better understand their surroundings, recognize objects, and adapt to changing conditions.

 

The training center follows a real-world simulation approach and currently includes five practical application scenarios:

• Guest reception and guidance

• Industrial cargo handling

• Household living environments

• Commercial retail operations

• Logistics and assembly activities

​

One of the key objectives is to improve robots' generalization capabilities. For example, if a robot is trained to move a stool and the stool is later replaced with a different model, the AI system can still recognize it as a stool and successfully complete the task. This ability to adapt to new objects and situations is considered essential for future large-scale deployment. The Xiong’an facility has also developed a unified platform that supports equipment management, data collection, data labeling, and model training for heterogeneous robots from multiple manufacturers. This allows different robot models to be trained using a common infrastructure and shared datasets.

 

While training centers provide the data foundation, industrial deployment is already underway. Chinese robotics company Robotera, based in Beijing, has deployed humanoid robots in more than ten logistics centers operated by China Post and SF Holding.

 

During recent demonstrations, robots successfully packed products into shipping boxes and automatically corrected mistakes when items were removed, illustrating the growing maturity of embodied AI systems.Industry experts view the government's latest initiative as a turning point for China's robotics sector. The focus is no longer on showcasing isolated technological achievements but on creating integrated systems capable of performing useful and repeatable tasks in real-world environments.

 

By combining large-scale training facilities such as the Xiong’an Embodied Intelligence Training Center, substantial government support, and ambitious deployment targets, China is positioning itself as one of the world's leading players in the commercialization of humanoid robots. The coming months will be critical in determining whether embodied AI can successfully transition from experimental demonstrations to widespread industrial adoption.