| license: mit | |
| language: | |
| - en | |
| tags: | |
| - robotics | |
| - simulation | |
| - openusd | |
| - mujoco | |
| - agilex-piper | |
| - dvidia | |
| pretty_name: DVIDIA USD workstation v0.1 | |
| size_categories: | |
| - n<1K | |
| # DVIDIA USD workstation v0.1 | |
| A bounded OpenUSD-to-MuJoCo workstation with a pinned PiPER arm and original rigid test geometry. All 45 passive physics cases are retained: 36 normal drop/slide checks and 9 declared controls that remove gravity, support or friction. These are physics checks, not a learned manipulation success rate. | |
| [Open the recorded workstation](https://dvidia-robot-simulation-lab.static.hf.space/usd-workstation/v01/index.html) · [Source and Terminal tools](https://github.com/Dvidia-Inference/dvidia-training) | |
| Download `usd-workstation-v01.zip`, extract it, and serve the directory with `python -m http.server 8000`. The browser replays saved native poses; the native Python CLI separately runs physics. See `guide.md` for installation and unsupported USD features. | |
| The read-only audit reconstructs 10 ms traces and saved body poses. It does not authenticate physics, reconstruct unrecorded 1 ms goal-monitor intervals, or establish calibrated shoe properties, a learned skill or physical robot readiness. | |
| Original DVIDIA code and scene: MIT; pinned PiPER assets: MIT with notices in the archive. Other dependencies retain their own licenses. DVIDIA research · [@iammrriver](https://x.com/iammrriver) · hello@dvidia.org. | |