SPACE 002 / LEARNED POLICY

Go1
locomotion.

A joystick-conditioned quadruped policy trained with smoke-gated PPO, evaluated locally, and opened in SimRig’s browser preview.

CONNECTINGgo1 / joystick-flat-terrain

SIMRIG · THREE.JS

chunking up meshes

CAM 01 / ORBIT

drag to orbit · scroll to zoom · right-drag to pan

episode—

step—

reward—

playback—

TaskFollow planar velocity and yaw commands over flat terrain.

Training8,192 parallel environments with local and remote smoke gates.

ArtifactDownloaded policy, metrics, configuration, logs, and checkpoints.