NVIDIA Isaac Sim 5.0 Updates

• Open and customizable: Isaac Sim 5.0 will be open and fully customizable on GitHub, allowing users to tailor simulations to specific needs, like synthetic data generation and Software-in-the-Loop testing.
• MobilityGen: This feature in Isaac Sim will enable high-quality motion data generation for training adaptable robots in diverse environments.
• Expanding synthetic data generation: The update will include tools for generating synthetic data in physical spaces for applications like robotics, autonomous vehicles, and smart spaces.
• Improved data format compatibility: World foundation models like NVIDIA Cosmos Transfer will enhance the photorealism of synthetic data for model training.
• Enhanced sensor simulation: Standardized interfaces will simplify integration for developers by providing consistent control of simulations from ROS 2.

NVIDIA Isaac Lab 2.2 Updates

• Benchmarking scripts for GR00T N1 models: The latest update includes benchmarking tools for the GR00T N1 models in Isaac Lab.
• Enhanced synthetic motion data generation: Improved tools for generating synthetic motion data for training models in Isaac Lab.
• Faster load times with Omniverse Fabric: NVIDIA Fabric will enhance load times and scalability by optimizing scene data handling.
• Improved suction cup gripper modeling: Users can now create and configure deformable surface grippers in Isaac Lab for more realistic simulations.

The new versions of NVIDIA Isaac Sim 5.0 and Isaac Lab 2.2 will be available this summer, offering advanced physics simulations, customization options, and benchmarking tools to accelerate robotics development.