Uranus: Building the Next-Generation Simulation Infrastructure for Embodied AI

Tue, 25 Aug 2026 00:00:00 +0000

Background: Why a learned simulator

“What I cannot create, I do not understand.” — Richard Feynman

To act intelligently, a robot must first imagine the consequences of its actions — yet real interaction is costly, slow, and impossible to reproduce, while traditional simulators (Isaac Sim, MuJoCo) demand hand-authored assets and physics and still suffer a sim-to-real gap. Uranus closes this gap with a data-driven world model: feed it observations, camera calibrations, joint states, and a robot description file, and it autoregressively generates multi-view video conditioned on actions — simulation learned from data, not hand-built.

About

Mon, 01 Jan 0001 00:00:00 +0000

We are the Large Model Team at D-Robotics, an AI research and engineering team building world models and scalable systems for intelligence in the physical world.

Our work connects generative modeling, robotics, and systems because physical intelligence depends on all three.

What we do

World Models — predictive and controllable representations of the physical world.
Generative Video — temporally coherent visual futures grounded in geometry and action.
Robot Learning — learning and evaluation across tasks, scenes, and embodiments.
AI Infrastructure — data, training, inference, and serving systems built to scale.

Connect

GitHub

D-Robotics Large Model Team

Uranus: Building the Next-Generation Simulation Infrastructure for Embodied AI

About

What we do

Connect