Home Bio-inspired Robots HuB: Humanoid Robots Learning Extreme Balance

HuB: Humanoid Robots Learning Extreme Balance

34
0
🤖 check out more robots like this

Credit: Tong Zhang, Boyuan Zheng, Ruiqian Nail, Yingdong Hu, Yen-Jen Wang, Geng Chen, Fanqi Lin, Jiongye Li, Chuye Hong, Koushil Sreenath, Yang Gao

Teaching robots how to maintain their balance takes work. We have already seen Optimus and other robots performing challenging moves. Some of them can even do forward and sideflips. HuB explores the idea of teaching humanoid robots extreme balance. It enables humanoids to perform quasi-static balance tasks with high stability.

HuB: Learning Extreme Humanoid Balance

With this approach, robots will be able to do deep squats, single leg stands, Bruce Lee’s kick, and high knees. As the researchers explain:

“HuB (Humanoid Balance) is a unified framework that integrates reference motion refinement, balance-aware policy learning, and sim-to-real robustness training, with each component targeting a specific challenge.”

[HT]

*Our articles may contain aff links. As an Amazon Associate we earn from qualifying purchases. Please read our disclaimer on how we fund this site.