Undergraduate researcher at Physical Intelligence Lab, ZJU, advised by Prof. Liangjing Yang and previously by Prof. Hua Chen. Research on embodied AI, focusing on legged locomotion and whole-body loco-manipulation. Recipient of Gold Medal & First Prize at the World Robot Contest Finals (WRCF) and Outstanding Student of Zhejiang University (2025–2026). Maintainer of ENotes, an open notebook on robotics, embodied AI, and university coursework, with research notes and personal logs under More (mostly in Chinese).
My research focuses on embodied AI, especially legged locomotion and whole-body loco-manipulation. I am also eager to explore other directions that move robotics forward, such as vision-language-action (VLA) models and world models.
Recently, I contributed to a whole-body controller for a bipedal mobile manipulator that tracks a single 6-DoF end-effector target, coordinating arm motion, posture adaptation, and stepping without explicit locomotion commands. My work covers:
Reward tuning for RL-based whole-body control in Isaac Lab
Sim2Sim validation of trained policies
Real-robot deployment on a LimX TRON 1 biped with an ARX L5 arm
I have also worked with UMI-based diffusion policy pipelines.
I come from a robotics-competition and mechanical-design background (WRCF, VEX, ZJU Mechanical Design Competition), which keeps me grounded in how things work on real hardware. I am eager to bridge the gap between simulation and reality to enable robots with more agile and dexterous behaviors.