NVIDIA Isaac Sim

NVIDIA Isaac Sim is an extensible robotics simulator enabling developers to design, simulate, test, and train AI-based robots through sensor and robot simulation, reinforcement learning, and synthetic data generation. It's powered by Omniverse to deliver scalable, physically accurate virtual environments for building high-fidelity simulations. Powerful built-in capabilities include NVIDIA Omniverse Replicator for generating synthetic data and Isaac Lab for GPU-accelerated reinforcement learning. Isaac Sim also includes accurate sensor simulations for most of the popular sensors available today. It supports ROS and ROS 2.

Developers can integrate Isaac Sim technologies as APIs into their existing applications or customize a version of the Isaac Sim reference application.

Isaac Sim communicates with MATLAB^® and Simulink^® via ROS messages. Isaac Sim, MATLAB, and Simulink all enable robotics simulation. A popular use case involves bringing synthetic sensor data from Isaac Sim into MATLAB and Simulink models. With the goal of accelerating the simulation of AI-based robots, Isaac Sim natively supports AI and reinforcement learning training and testing with software-in-the-loop (SIL) and hardware-in-the-loop(HIL). Finally, a program developed in MATLAB and Simulink can be deployed on embedded hardware (for example, on the NVIDIA Jetson platform) to drive the end application. For more information on connecting Isaac Sim with MathWorks’ ROS Toolbox see the Setup documentation and this introductory webinar.