RealMan Robotics has launched a Humanoid Robotics Data Training Center in Beijing to advance the development and application of humanoid robots. The facility, covering 3,000 square meters, is designed as a hub for technology research, scenario-based application testing, operator training, and ecosystem collaboration. RealMan, a robotics technology and equipment provider, is responsible for the center’s deployment and daily operations.
The center is equipped with 108 robots, including dual-arm lifting models, wheeled humanoids, drone-arms, and quadruped platforms. To generate training data under realistic conditions, it has constructed ten environments simulating use cases such as eldercare, rehabilitation, special operations, retail, automotive assembly, and catering. According to the company, these scenarios will produce more than one million multimodal data points annually to support artificial intelligence model training.
The initiative seeks to address three obstacles to robotics development: insufficient data generalization across scenarios, discrepancies between simulation and real-world conditions, and a lack of standardized data formats for efficient iteration. The facility’s full-stack pipeline, covering collection, training, validation, and deployment, is intended to support the commercialization of humanoid robotics and embodied AI.
At the center’s Open Day, Eric Zheng, Director of the Humanoid Robotics Data Training Center, identified operational capability, generalization, and cost efficiency as major challenges hindering wider adoption of robotics. He noted that existing industrial and service robots remain limited in adaptability and affordability, slowing their integration into complex real-world environments.
In conjunction with the launch, RealMan introduced the RealBOT Embodied Intelligence Open Platform, aimed at enabling high-quality data acquisition through integration with remote teleoperation systems. The company described the platform as a step toward advancing human-robot collaboration and improving the scalability of robotics applications. The center plans to expand partnerships with industry and academic institutions, promote data sharing, and foster ecosystem collaboration to accelerate the adoption of humanoid robotics globally.
