- Published on
Xiaomi's CyberOne 02 clocks an 8-hour factory shift as Unitree eyes a robotics inflection point
- Authors
- Name
- DeepHorus
Lei Jun recently released a video of Xiaomi's humanoid robot working in a factory, showing CyberOne 02 completing the sorting of center console side covers in Xiaomi Auto's final assembly workshop. The video shows that the robot accumulated 8 hours and 5 minutes of work. Lei Jun used this to disclose new progress in the Xiaomi humanoid robot's "internship."
Center console side covers are relatively large, irregularly shaped flexible workpieces. The robot needs to follow the vehicle model line-up sequence, take the cover out of the raw material box, adjust its posture, and then place it into the corresponding rack slot. In the video, two humanoid robots operate according to the production line takt and coordinate with AMRs transporting materials to adjust their material-picking routes and stride lengths.
During the test, the robots charged after accumulating 3 hours and 2 minutes, and 6 hours and 3 minutes of work, respectively. In the latter half, one robot dropped a part, which was picked up by a wheeled robot. These details show that humanoid robots still need to address battery life and exception-handling issues in real production lines.
Regarding why humanoid robots are used in factories, Lei Jun responded that the hardware of a general-purpose humanoid robot does not need to be modified and can adapt to more positions and scenarios. What Xiaomi demonstrated is a specific process in the final assembly workshop, not a single demonstration action. The test also makes "general-purpose hardware + multi-position adaptability" the main selling point for humanoid robots entering factories.
At the World Internet Conference Wuzhen Summit on November 4, Wang Xingxing, founder of Unitree Robotics, said that next year will most likely be a critical point for the robotics industry or robot models. He previously said at the China International Industry Fair on October 12 that Unitree has been conducting pilot deployment with automobile companies since 2024. However, moving from single-point experiments to large-scale application still requires improving generalization capabilities and operational efficiency.
Wang Xingxing believes that some enterprise-oriented commercial service scenarios may advance faster, while home services have higher requirements for adapting to complex environments. In terms of computing power allocation, he proposed keeping computations with high real-time requirements, such as motion control, on the robot body, while relying on the cloud as much as possible for complex decision-making and adaptation to large-scale scenarios. Xiaomi's factory test echoes industry judgments: humanoid robots are moving from demonstrations to process verification, but battery life, reliability, and efficiency remain questions that must be answered before large-scale deployment.
