High-performance power joint modules available in 10, 30, 60, 120, 240, and 360 N·m rated torque options, with electromagnetic brakes, dual encoders, and a hollow-shaft design for multi-axis integration.
Primary use case: Multi-axis robotics integration
China is rapidly expanding its humanoid robot innovation centers, with 22 provincial-level centers and at least 90 lower-tier data collection sites now operating, under construction, or in planning. The Wuhan center alone generates 24,000 training data entries daily, with government subsidies covering operations even when robots fail, as China seeks to secure vast troves of physical interaction data to gain an advantage in the AI technology competition with the United States.
RealMan announced the launch of its global link network robot deployment initiative at the World Robot Conference 2026. The initiative appears to focus on expanding robot deployment capabilities internationally through a networked approach.
Robbyant, an embodied AI company within Ant Group, has announced the upgrade and open-source release of LingBot-VLA 2.0, a next-generation vision-language-action model for robotics pre-trained on 60,000 hours of real-world physical data from 20 robot morphologies. The model demonstrates superior cross-morphology generalization, outperforming π0.5 and GR00T N1.7 on dual-arm manipulation benchmarks and achieving 3x inference efficiency improvement with latency under 150ms. Robbyant is currently conducting commercial pilots with hardware partners Leju and Ti5Robot, as well as enterprise customers GuoDa Drugstore and Longsheng Technology in retail, logistics, and industrial environments.