| tags: | |
| - microduck | |
| - robotics | |
| - reinforcement-learning | |
| - onnx | |
| library_name: microduck | |
| pipeline_tag: robotics | |
| # backroll | |
| Backward roll: rolls over the back and stands up. Sim-only, not yet tested on a real robot - try on a soft surface first. | |
| A **episodic** policy for the [microduck](https://github.com/pollen-robotics/microduck) (61-D observation, 14 actions, 50 Hz). Runs 4.0 s and returns itself to a standing pose. | |
| ## Run it on a robot | |
| ```bash | |
| sudo robotctl policy add backroll IgaPro/microduck-backroll | |
| robotctl robot do backroll | |
| ``` | |
| The observation normalizer is baked into `policy.onnx`; feed raw observations. | |
| `manifest.json` follows schema 2 of the microduck policy manifest (`docs/policy-manifest.md` in the daemon repo). | |
| ## Training | |
| - **task_id**: `Mjlab-Backroll-Flat-MicroDuck` | |
| - **repo**: `pollen-robotics/microduck_rl` | |
| - **branch**: `backroll` | |
| - **commit**: `b02fdea49` | |
| - exported from a checkout with uncommitted changes | |