Skip to content
RobothecaBeta
Log in

LearningHumanoidWalking

  • Software
  • Humanoids and robot learning
  • Framework

Code for research on training humanoid robots to walk with deep reinforcement learning, covering uneven and compliant terrain, planned footsteps, and current feedback.

Source on GitHub: github.com/rohanpsingh/LearningHumanoidWalking (opens in a new tab). Facts read at commit cd8c655 (opens in a new tab).

Licenses

Code
BSD-2-Clause (opens in a new tab)

Read by Robotheca from LICENSE.md (opens in a new tab) (BSD-2-Clause) at commit cd8c655 (opens in a new tab), on 22 Oct 2026.

Commercial use
Yes

What it does

  • Trains bipedal walking policies for humanoid robots using reinforcement learning, with a simulation base class for environments.
  • Supports environments for a Unitree H1 standing task, JVRC walking and stepping tasks, and a cartpole swing-up test task.
  • Walking can be learned on uneven or compliant ground, following planned footsteps, and with simulated back-EMF current feedback.
  • Environment behavior is set through YAML configuration files covering timestep, gait phase durations and reward weights.
  • Adding a new robot means creating an environment class that inherits from the base humanoid environment and registering it.
  • Training and evaluation are run through a single run_experiment.py script.

Written by Robotheca from the project's README (opens in a new tab) at commit cd8c655.

Getting started

Difficulty
Advanced

Maintenance

Status
Maintained: a commit in the last 12 months
Newest commit
2 May 2026 on main
Latest release
v1.0.0 (25 Feb 2025)

Checked on GitHub on 22 Oct 2026.

Links

Robots that can run it

No robot on Robotheca is linked to it yet. Links come from a robot's robot.yaml, from its dependency files read as text, or from Robotheca's editors.

Robotheca lists software as a reference: nothing is hosted or run here, and every fact links to where it was read.