Skip to content
RobothecaBeta
Log in

Asimov 1 locomotion policy

  • AI model
  • Humanoids and robot learning
  • Runs on hardware you can build

Training and evaluation code for Asimov 1 humanoid locomotion policies, built as an Isaac Lab extension using PPO and adversarial motion priors, for sim-to-real deployment.

  • Closed core

Made by Menlo Research, Singapore (checked on github.com/menloresearch (opens in a new tab)).

Source on GitHub: github.com/menloresearch/cyclotron (opens in a new tab). Facts read at commit 870c161 (opens in a new tab).

Licenses

Code
BSD-3-Clause (opens in a new tab)

Read by Robotheca from LICENSE (opens in a new tab) (BSD-3-Clause) at commit 870c161 (opens in a new tab), on 16 Oct 2026.

Commercial use
Unknown

It needs closed parts under their own terms: read them before any commercial use.

One part it needs is closed

  • Needs NVIDIA Isaac Sim components (training): Training goes through Isaac Lab, which needs Isaac Sim's closed components; the policy itself runs without them.

Labels

  • Closed coreIt needs closed parts: NVIDIA Isaac Sim components (training).

What it does

  • Trains Asimov 1 walking policies in Isaac Lab with PPO, with an optional adversarial motion prior (AMP) variant.
  • Plays back checkpoints in Isaac Sim to watch and evaluate trained policies.
  • Exports checkpoints to ONNX with a deployment contract in the metadata for robot runtimes.
  • Uploads trained runs with a generated model card to the Hugging Face Hub.
  • Views exported or shared policies in the browser with MuJoCo, without Isaac Sim or a GPU.
  • Includes a quick install script that sets up uv, Isaac Lab and the Asimov 1 robot model.

Written by Robotheca from the project's README (opens in a new tab) at commit 870c161.

Getting started

Difficulty
Advanced
Hardware it needs, as the project states it
Ubuntu 22.04+ on x86_64 with a compatible NVIDIA driver and a GPU; tested on RTX A6000, RTX PRO 6000, RTX 4090 and RTX 3090. Training at 4096 environments is said to suit A6000 or PRO 6000, with lower num_envs suggested for out-of-memory errors. The browser viewer needs only Node.js 20 or newer, wit

Maintenance

Status
Active: a commit in the last 90 days
Newest commit
16 Oct 2026 on main
Latest release
None published on GitHub

Checked on GitHub on 16 Oct 2026.

Links

Robots on Robotheca that use 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 AI models as a reference: nothing is hosted or run here, and every fact links to where it was read.