Qiyuan Robot

Headlines · 3

1. 3d
   Qiyuan Robot Integrates With Tencent WorkBuddy

   At the 2026 Tencent Global Digital Ecosystem Conference, Qiyuan Robot CEO Tian Hua shared the latest progress on integrating the robot with Tencent WorkBuddy. According to Tian Hua, users can issue natural-language instructions that WorkBuddy interprets and breaks into tasks, which Qiyuan Robot then executes using environmental perception and action capabilities.

   36氪快讯 · 🔥 0
2. 4d
   Cars emerge as an early commercial vehicle for embodied AI

   A TMTPost analysis credited to 科技新知 contrasts reports of Unitree Robotics’ 16% share-price rebound and RMB200 billion market capitalization with DeepBlueRobot’s unpaid wages and court-ordered liquidation. It argues that automakers’ mature supply chains, large volumes of real-world driving data and proven engineering systems make cars a practical starting point for validating physical AI. The analysis cites work by Li Auto, XPeng, AIVA and Qianli Technology, including models such as Li Auto’s L9 Livis and AIVA’s M7. It also says that transferring automotive capabilities to more general humanoid tasks remains difficult because robots additionally require motion control, dexterous manipulation, world models, spatial reasoning and manufacturing consistency.

   钛媒体 · 🔥 0
3. Oct 20
   Peking University and Qiyuan Robot unveil HIL-UMI for arm-free robot post-training

   A Peking University team led by Dong Hao, who is also chief scientist at Qiyuan Robot, has introduced HIL-UMI, a human-in-the-loop data collection method for embodied post-training that runs without a robot arm. An operator uses a handheld UMI gripper while a model predicts the next actions in the background; when its predictions diverge from the human demonstration, the system prompts the operator to record corrective data, using an energy score and an advantage estimator. In four real-arm tasks, the team reports an average task-progress score of 94.8 with HIL-UMI versus 58.0 for conventional supervised fine-tuning, and 73.40 ms versus 412.99 ms per effective frame on a desk-cleaning task, a 5.63x collection-efficiency gain.

   雷峰网 AI · 🔥 0

Experience and discussion from the community

Share my Qiyuan Robot experienceAsk about Qiyuan Robot