This site requires Javascript in order to view all its content. Please enable Javascript in order to access all the functionality of this web site. Here are the instructions how to enable JavaScript in your web browser.

Learning Path

Digital Twins for Physical AI

Get started with essential foundations in developing OpenUSD-based digital twin applications and workflows for the era of physical AI.

Introduction
Start Learning
Next Step

Introduction
Start Learning
Next Step

Accelerate your digital twin development journey through this free NVIDIA Deep Learning Institute (DLI) learning path. Explore foundational OpenUSD concepts, learn how to develop with NVIDIA Omniverse™ SDKs and libraries, and dive into the fundamentals of synthetic data generation, world foundation models, and robotics.

Start Your Free Learning Path

Choose the Training That Best Matches Your Needs

Click on the tiles below to launch.

Digital Twin Foundations

Learn about digital twins, why they matter for physical AI, and how Omniverse technologies and OpenUSD are enabling the development of custom digital twin applications.

What Is a Digital Twin?

Explore real-world examples of how digital twins are used to simulate, optimize, and visualize physical systems and environments using technologies like OpenUSD, RTX™ rendering, and generative AI.

Fundamentals of Working With OpenUSD

Start here to discover how USD layers help with scene composition and non-destructive workflows, and how data separation and asset reuse accelerate 3D workflows.

Learn OpenUSD: Foundations Learning Path

Once you understand how USD is used in simple digital twin scenes, complete this series to learn more essential concepts, vocabulary, and Python best practices.

Getting Started With Omniverse Development

Develop your own digital twin solutions using Omniverse, a platform of APIs, SDKs, and services that enable developers to integrate OpenUSD, NVIDIA RTX, and generative physical AI into existing software tools and simulation workflows.

An Introduction to Developing With NVIDIA Omniverse

Build custom applications with NVIDIA Omniverse Kit SDK and templates.

Creating an Omniverse Extension With Python

Learn the basics of Omniverse extension development in this course, where you’ll write Python code to build a simple, interactive extension.

Extend Omniverse Kit Applications for Building Digital Twins

Learn how to manipulate OpenUSD scenes by modifying USDA files, and customize the USD Explorer template application by adding physics using Python.

Customizing UI and Front-End Dashboards for Digital Twins

Make your application unique through customized UI and front-end experiences for your audience.

Coming soon.

Building and Deploying Digital Twin Applications With Omniverse Kit App Streaming

A hands-on guide to creating, deploying, and extending 3D digital twin applications using NVIDIA Omniverse.

Assembling OpenUSD Data for Digital Twins

Learn how to harness the extensible framework of Universal Scene Description (OpenUSD) to assemble diverse datasets for building and simulating in virtual 3D worlds.

Assembling Your First Digital Twin With Omniverse and OpenUSD

Assemble, organize, and prepare factory assets and scenes using OpenUSD and Omniverse to create modular, simulation-ready digital twins for manufacturing and physical AI.

Preparing 3D Assets for Simulation and Physical AI

Discover how 3D assets are prepared for scalable simulation and machine learning applications.

Coming soon.

Learn OpenUSD: Creating Composition Arcs

Learn how USD composition arcs like layers, references, payloads, and variant sets are used in complex 3D scenes.

Learn OpenUSD: Asset Structure Principles and Content Aggregation

Explore key asset structure concepts in OpenUSD to aggregate, organize, and scale modular 3D scenes.

Learn OpenUSD: Developing Data Exchange Pipelines

Extract, transform, and validate 3D data from your existing CAD tools into OpenUSD-based scenes.

Learn OpenUSD: Asset Modularity and Instancing

Leverage asset modularity, instancing, and content reuse to build efficient 3D scenes at scale.

Coming soon.

Digital Twins for Simulation and Physical AI

Learn how to use digital twins to train and optimize physical AI systems, from generating and curating synthetic data for robotics with world foundation models to enabling agentic workflows for autonomous machines.

Synthetic Data Generation With Generative AI

Use NVIDIA APIs like USD Code to automate and enhance synthetic data pipelines with generative AI.

Robotics Fundamentals Learning Path

In this structured learning path, learn how you can use Isaac Sim™ and digital twins to simulate, train, and test robots for manufacturing and industrial tasks.

An Introduction to NVIDIA Cosmos for Physical AI

Experiment with Cosmos™ World Foundation Models (WFMs) to generate and curate physically accurate synthetic data for training physical AI.

Building AI Agents With Multimodal Models

Build neural network agents that reason across multiple data types for real-world applications leveraging video search and summarization.

Coming soon.

This path is a suggestion and might be modified based on skill levels and desired outcomes.
© 2025 NVIDIA Corporation and affiliates. All rights reserved. NVIDIA constantly develops new courses and updates current content. Courses may be added and/or removed. Hyperlinks may change.

OpenUSD certification badge

Ready to Get USD Certified?

Take the next step in your 3D development career with OpenUSD certification.

Additional Resources

Build Digital Twin Solutions With NVIDIA’s Developer Templates

Accenture, KION Group

Test Multi-Robot Fleets for Industrial Automation

Simulate and test physical AI and robotic fleets at scale in industrial digital twins before real-world deployment.

Microsoft Azure, Rockwell Automation

Build Digital Twins for Remote Monitoring

Enable remote monitoring of industrial operations with accurate digital twins integrated with real-time IoT data.

Explore Workflow

Luminary Cloud

Build a Digital Twin for Interactive Fluid Simulation

Integrate real-time visualization and analysis into your 3D workflow to enable real-time computer-aided engineering of your product.

Bootstrap Synthetic Data Generation With Generative AI

Build an industrial digital twin and apply generative AI to enhance data creation for perception AI model training.

Explore Workflow

Foxconn

Spatial Streaming for Omniverse Digital Twins

Stream immersive OpenUSD-based Omniverse digital twins to the Apple Vision Pro with this reference workflow.

Explore Workflow

Cadence, ETAP, Schneider Electric, Vertiv

Build Digital Twins for AI Factory Design and Operations

Design, test, and optimize a new generation of intelligence manufacturing data centers using digital twins.

Interested in a particular topic? View our full set of DLI learning paths to see the recommended sequence of courses to take.

View All Learning Paths