Xiaomi-Robotics-1: Scaling Vision-Language-Action Models with over 100K Hours of Real-World Trajectories
Core Contributor: Qiao Sun · Xiaomi Robotics Team
Scaling robotic learning with over 100K hours of real-world manipulation trajectories.
Qiao SUN
Xiaomi Robotics · Robot Foundation Model Team
Towards Building Embodied Intelligence that Can Understand & Interact with the Physical World.
I am a Researcher with the Robot Foundation Model Team at Xiaomi Robotics, after conducting visiting research at UMass Amherst & MIT-IBM Watson AI Lab under Prof. Chuang Gan and Dr. Yilun Du. My academic journey took an unconventional path — from Civil Engineering and Financial Management at Tianjin University, to Electrical and Computer Engineering at Fudan University, before fully committing to AI research.
My research centers on embodied AGI — building robots that can perceive, reason about, and physically interact with the world. I work at the intersection of world models, vision-language-action models, and scalable robotic learning, aiming to translate the recent breakthroughs in foundation models into the physical world.
