PhysicalAI WorldModel Synthetic Embodied Robot Scenes Dataset Card Dataset Description PhysicalAI WorldModel Synthetic Embodied Robot Scenes is a large scale synthetic robotics video corpus generated from USD based robotic simulation and rendering pipelines built around NVIDIA Isaac Sim, Omniverse, Isaac Lab, and related robot data generation systems. It is designed to improve physical plausibility, embodiment persistence, task conditioned robot behavior reasoning, contact understanding, and long horizon robot video modeling for Cosmos training. For simplicity, this card abbreviates the dataset as SDG RobotSim in some sections. The v1.0 release contains 373,703 RGB MP4 video clips across three top level task families: collision, manipulation, and humanoid motion. The release covers mobile robots, quadrupeds, humanoids, fixed base manipulators, bimanual systems, and dexterous hand arm embodiments. Rather than modeling one robot platform exhaustively, SDG RobotSim exposes models to a broad range of robot embodiments, rendered environments, camera viewpoints, task conditions, and contact rich behaviors. The dataset combines simulated robot rollouts, policy rollouts, retargeted humanoi…
Runs entirely in your browser via DuckDB-Wasm — this dataset's real data file is loaded once, then queried locally. Nothing is sent to a server.
We use cookies for essential functionality and analytics. You can accept or reject analytics cookies.Cookie policy