Skip to content
humanoidrobots.training

Starting viewport

PAL Robotics · humanoid

Kangaroo

Research biped/humanoid with custom linear actuators and configurable arms (14-40 DOF), 1.6 m and 50-65 kg.

Frames · Kangaroo

2 items · credit and license on each
  • PAL Robotics | Kangaroo ProPALRobotics · official embed · source
  • PAL Robotics | KANGAROO Free Walking DemoPALRobotics · official embed · source

Provenance

Model
PAL Kangaroo
Entry
kangaroo_mujoco/models/mjcf_data_floating_4dof_no-end-effector_no-end-effector_fixed/scene_empty.xml
Files
26 · 28 MB, fetched from raw.githubusercontent.com, cached in your browser
License
Apache-2.0
Engine
MuJoCo 3.14.0, official WebAssembly build · three.js renderer
Controller
PD hold on the stand keyframe with an ankle-strategy tilt term; not a learned policy
Note
PAL's own MJCF (kangaroo_simulation, branch humble-devel), 4-DoF arms. Its assets folder is a git symlink that raw GitHub does not follow, so the tab points meshdir at the real folder.

Policies listed for Kangaroo

No published policy we can replay yet.

Nothing is playing a trained controller in the viewport above. What you see is a PD hold on the stand pose, computed in your tab. A policy plays here when a checkpoint is published with an obs/action contract we can reproduce exactly in MuJoCo, and it passes a 30 s walk test in the tab.

Set up a training run

Train on GPUs

mjlab ships no stock task for Kangaroo. The training page gives you the pinned model and the steps to write a robot config.