Donny Li of RealMan discusses how robot hardware and real-world training data can accelerate everyday robot use at the Humanoids Summit.
Li says the goal is to put robots to work in settings such as retail, food service, and healthcare. He argues that progress depends on collaboration because no single company or country can build the full stack alone.
Li identifies robot bodies, algorithms, AI models, computing, and data as key parts of embodied intelligence. RealMan concentrates on robot hardware and real-robot data, including lightweight arms designed for practical deployment.
Li points to data quality, breadth of offerings, and production volume as reasons to use RealMan’s real-robot data. The company uses synchronization checks and other quality controls, and its Bino version adds two head cameras for a binocular view.
RealMan has deployed 108 robot sets across more than 10 environments, with teleoperation and different grippers or accessories to suit collection needs. Li says customers should consider standard datasets before commissioning custom work, since at least half of those seeking specialized data found the standard offering sufficient.
Li reports daily capacity of 60,000 data entries. He also says RealMan released 130 hours of video data publicly and received more than 10,000 downloads in two weeks; the accompanying facility video shows tasks including folding towels, handling fragile food, and agricultural harvesting.
Briefs summarise talks and interviews; timestamps open the video at that moment. Claims in a talk are the speaker's own.
Edward Liu of RealMan Robotics explains how RealMan is building hardware, fleet operations, and data infrastructure to move physical AI from demos to daily work at the Humanoids Summit.
Kenneth Wren, director of overseas solutions at RealMan Robotics, discusses robot training and human-robot collaboration in a Bloomberg interview.
Jenk Infang, head of RealMan Robotics’ humanoid robot data training center, introduces its Beijing facility and explains how it collects data to train humanoid robots for household and workplace tasks.
