Reinforcement learning enables quadruped robots to jump on the moon

End-to-End Reinforcement Learning for Quadrupedal Jumping in Lunar Gravity

Robotics

Summary

Robots that walk on legs can explore difficult places on the moon, but sometimes they need to jump to get over big gaps or obstacles. The authors used a special kind of computer learning called reinforcement learning to teach a four-legged robot how to jump from one spot to another in moon-like gravity. The robot learns to jump accurately and land safely using only sensors it could realistically carry on a real mission. They tested the robot in computer simulations and on real hardware, and the jumps were both precise and feasible for the robot's motors and power.

What this means in practice

  • For robotics engineers: Deploy control policies that enable legged robots to jump across terrain gaps in low-gravity environments like the moon.
  • For space mission planners: Design robotic explorers with advanced mobility for lunar missions that include jumping to cross obstacles and rough terrain.

Authors

Levin Müller, Oliver Fischer, Marco Trentini, Yusuke Tanaka, Marco Hutter

Abstract

Legged robots can extend planetary exploration to rugged and discontinuous terrain that is inaccessible to conventional rovers. In lunar gravity, jumping further expands this mobility by enabling robots to cross obstacles and gaps that cannot be traversed by walking alone. We present a multistage reinforcement-learning approach that produces a single end-to-end control policy for commanded point-to-point jumping under lunar gravity. The resulting policy has a simple architecture and relies exclusively on mission-compatible onboard sensing, making its deployment on a real lunar mission plausible. In simulation, the learned policy performs accurate point-to-point jumps with landing errors smaller than the robot's body size. We analyze the torque and power demands of lunar jumping and characterize performance degradation under extreme outlier events, including control-update freezes of up to \SI{1}{s} during flight. Hardware transfer demonstrates that the policy can successfully execute a vertical jump under lunar-gravity offloading. Furthermore, the measured torque and power profiles are consistent with the simulation results and remain within the robot's hardware limits.