Uranus: Building the Next-Generation Simulation Infrastructure for Embodied AI
Uranus is a diffusion robot simulator that streams open-ended embodied rollouts at 24 FPS.
Uranus is a data-driven robot simulator built on a joint-trajectory-conditioned autoregressive diffusion model for embodied AI. It accepts future joint-position trajectories online and generates one latent frame per step, equal to four RGB frames, without a fixed horizon. Optimized inference reaches 24 FPS, and a unified interface supports synchronized multi-view generation across robot embodiments and cameras. The authors evaluate in- and out-of-distribution data and release the code and weights.
- Joint-trajectory-conditioned autoregressive diffusion generates one latent frame per step.
- Streaming rollout has no fixed horizon and runs at 24 FPS.
- Unified interface supports synchronized multi-view generation across robot embodiments.
- Code and model weights are released after in- and out-of-distribution tests.
Full article142 words · extracted from huggingface.co · click to collapse
Scalable simulation is essential for robot data generation, policy training, evaluation, and safe iteration, yet real-world interaction is costly and conventional simulators require labor-intensive construction. We present Uranus, a data-driven robot simulator built around a joint-trajectory-conditioned autoregressive diffusion model. Uranus offers three key capabilities: (1) streaming, open-ended rollout, which receives future joint-position trajectories online and autoregressively generates one latent frame per step, corresponding to four RGB frames, without a fixed horizon; (2) low-latency generation, achieving 24 FPS after inference optimization; and (3) scalable, extensible robot control, providing a unified interface for synchronized multi-view generation across diverse robot embodiments and camera configurations. We conduct comprehensive quantitative and qualitative evaluations on both in-distribution and out-of-distribution data, providing an objective assessment of Uranus and clearly identifying its current limitations. We release the code and model weights to empower the community with practical tools and insights.
Text extracted automatically; images, tables and formatting may be missing. Original: https://huggingface.co/papers/2609.24815