Data Regulator Chairs Sept 10 Embodied AI Symposium; Factory Tests Advance
On September 10, Liu Liehong, director of China's National Data Administration, chaired a symposium on embodied intelligence attended by research institutes and companies including Beijing Humanoid, Lightwheel and Galaxea AI. In the same period, PNDbotics demonstrated humanoid robots performing manufacturing tasks in a FutureFactory scenario at the Ningbo Smart City Expo.
The symposium places the data regulator, not only science and industry ministries, at the center of embodied intelligence policy. Data rules, annotation standards and cross-enterprise data circulation are likely to become the binding constraints for model training. The BRICS New Delhi Declaration separately praised its smart manufacturing and robotics working groups and endorsed AI governance cooperation, giving Chinese robotics firms a multilateral channel that can lower market-entry friction abroad.
PNDbotics' factory demonstration is a useful marker: real production tasks, even in an expo setting, indicate that deployment has moved beyond scripted stage shows. The value lies less in the robot than in the operational data generated, which feeds back into model iteration. Against that benchmark, most humanoid demos still lack hour counts, task success rates and cost-per-unit data.
Capital continues to fund early-stage bets. Xieyue Intelligence closed a several-hundred-million-yuan Angel+ round led by four institutions, with proceeds earmarked for embodied models and home robots. Its home-scenario focus differs from the factory-first positioning of most peers, suggesting investors are hedging across application tracks rather than converging on one.
Galbot's public denial of online rumors and its police report show that information risk is now a governance issue for leading players. Two observation points follow: whether the data administration issues concrete rules for embodied intelligence data, and whether factory deployments publish measurable operating data to justify the next funding cycle.