NVIDIA SimReady Central OpenUSD Asset Library Release

Overview of SimReady Central

SimReady Central is an open library hosted on Hugging Face that provides over 2,000 simulation-ready (SimReady) OpenUSD assets. These assets are specifically prepared for robotics and AI factory digital twin workflows, including pre-defined properties such as collision geometry, mass, friction, and thermal metadata. By providing standardized, ready-to-use models, the library aims to reduce the manual effort required to prepare CAD models for high-fidelity simulation environments.

What Defines a SimReady Asset

SimReady assets are OpenUSD files that contain the necessary physical and semantic metadata required for simulation runtimes. While standard CAD models provide geometry, SimReady assets include additional data layers required for functional simulation:

  • Collision Geometry and Physical Properties: Enables accurate contact and physics simulation.
  • Joint Structure and Limits: Defines how articulated equipment or robot arms move.
  • Semantic Labels: Supports object identification and the generation of annotated synthetic data.
  • Materials and Sensor Attributes: Facilitates rendering and sensor interaction modeling.

The SimReady Foundation provides the specifications, profiles, and validation tools used to ensure assets meet these requirements. Developers can select assets based on specific profiles that match their intended use case.

Capabilities for Developers and OEMs

SimReady Central serves both developers building simulations and equipment manufacturers (OEMs) looking to distribute their products in a simulation-ready format.

  • For Developers: The library supports natural-language search via NVIDIA USD Search, allowing users to query for specific equipment like robot arms, grippers, or warehouse infrastructure. These assets can be integrated into environments using tools like Isaac Sim and Isaac Lab.
  • For OEMs: Manufacturers can validate their CAD models against SimReady profiles and submit them to the library. This makes their equipment discoverable to developers, facilitating faster testing and decision-making for end customers. Participating OEMs include companies such as Fanuc, Comau, Franka Robotics, and Vertiv.

AI-Assisted Scene Authoring

Developers can utilize frontier AI models, such as GPT-6 Astra, to automate the construction of simulation scenes. The workflow involves using natural language to request specific scenarios—such as a robotic pick-and-place task or an AI factory layout—and having the AI agent perform the following steps:

  1. Query the SimReady Search API to locate appropriate assets.
  2. Reference or payload the assets into an OpenUSD stage.
  3. Generate the Python code required to assemble the workcell and configure tasks.
  4. Render previews for user review and refinement.

This approach allows for rapid iteration, such as comparing multiple factory floor layouts by checking for equipment overlaps and aisle-clearance violations against specified constraints.

Sources