Runpei Dong
Papers
1
Total Citations
4
H-Index
1
About
Runpei Dong is a robotics researcher whose work bridges the gap between simulated control and real-world humanoid dexterity. His primary focus lies in developing robust locomotion and recovery policies for legged robots, with a particular emphasis on enabling humanoid platforms to autonomously recover from falls. Dong’s most notable contribution, detailed in his highly cited 2025 paper “Learning Getting-Up Policies for Real-World Humanoid Robots,” introduces a novel reinforcement learning framework that allows a G1 humanoid robot to reliably stand up from both face-up and face-down positions. This work is groundbreaking for its demonstrated robustness across diverse, challenging terrains—including flat floors, deformable surfaces, slippery ground, and even sloped grassy fields—showcasing a level of adaptability previously unseen in real-world humanoid recovery. With 4 citations in a short time, this paper has already sparked interest for its practical implications in disaster response and household robotics. Dong’s achievements highlight a critical step toward truly autonomous humanoid robots capable of operating in unstructured environments without human intervention, making him a rising figure in the field of dynamic robot control and embodied AI.
Research Focus
Key Achievements
Top Papers
- 1Learning Getting-Up Policies for Real-World Humanoid Robots4 citations · 2025