pipette
ESEspañol

Solving a Giant Rubik's Cube with a Quadruped Robot

Lukas Amiet, Gabriel S\'anchez-Blanco Repgen, Mayank Mittal, Victor Klemm, Andrei Cramariuc, Marco Hutter, Cesar Cadena

Preprint

In the authors' words

Pedipulating a freestanding articulated object requires controlling its pose and internal configuration while maintaining balance. We demonstrate this capability by autonomously solving a giant Rubik's Cube with a quadruped using only its feet. The task demands repeated precise interactions, since unintended layer motion can invalidate subsequent operations. We develop an onboard RGB-D perception pipeline that estimates the cube's metric pose and its full color state from partial views acquired through locomotion and flipping. A planner converts the computed solve into robot-executable skill sequences. Learned whole-body skills reorient the cube while preserving its layer configuration and turn its layers with one foot pinning the cube as the other pushes. We train these skills using reinforcement learning with privileged observations and distill them into policies that map onboard depth and proprioception to joint targets. The deployed policies tolerate variations in relative robot-cube placement and recover from external disturbances. The system completes three autonomous hardware solves, each about 30 steps over 8-9 minutes. These results show that combining onboard perception with coordinated pedipulation enables a quadruped to complete extended multi-step physical tasks. Additional material can be found at: https://leggedrobotics.github.io/rubik-skills/

Main resultLimitation the authors admit

In plain words AI summary

Researchers built a system that lets a four-legged robot solve a giant Rubik's Cube using only its feet. The robot uses onboard cameras to estimate the cube's position and colors from partial views while moving around it. Learned skills let the robot flip the cube to reorient it and turn layers by pinning with one foot and pushing with the other. The system completed three autonomous solves on real hardware, each taking about 30 steps over 8-9 minutes. The results suggest that combining onboard perception with foot-based manipulation enables quadrupeds to complete extended multi-step physical tasks.

Written by Kimi K2.5 using only the abstract above. Jev checked that every sentence is supported by the abstract sentences it cites, and removed any that were not. Point at a sentence to highlight its source.

Appeared: Tuesday, October 20. arXiv. Preprint, not yet peer-reviewed.

Authors' comment: L. Amiet and G. S\'anchez-Blanco Repgen contributed equally. Project page: https://leggedrobotics.github.io/rubik-skills/