Build a Household Robot Workflow With NVIDIA Isaac and Cosmos
Join NVIDIA Omniverse Ambassador Kimate Richards, Founder of 10Things, and Oliver Choy, COO of 10Things, for a practical look at how one real robot uses the NVIDIA robotics stack. They’ll show how NVIDIA Isaac Sim supports robot development and testing, how Apple Vision Pro is used for teleoperation and data collection, and how Isaac Lab, GR00T, and Cosmos help with training, reasoning, and evaluation. They’ll also cover the robot’s onboard stack, including Jetson Thor running Isaac ROS for navigation, perception, and motion planning. The session will include Isaac Assist, an Omniverse extension built by the 10Things team to improve the simulation workflow, plus lessons learned from moving between simulation and real household robot tasks. What You’ll Learn - How Isaac Sim is used to develop and test a real household robot - How Apple Vision Pro supports teleoperation and data collection - How Isaac Lab, GR00T, and Cosmos fit into robot training and evaluation - How Jetson Thor and Isaac ROS support onboard robot autonomy - What worked, what did not, and what changed when moving from simulation to physical household tasks - How Isaac Assist helps streamline the simulation workflow ---------------------------------------------------------------------------- Chapters: 0:00 NVIDIA Isaac and Cosmos Household Robot Workflow 2:25 10 Things and Ada Robot Overview 10:01 CAD-to-Production Robot Workflow 13:26 Ada Elder Care Robot Architecture 18:15 Jetson Thor, Isaac ROS, and NVIDIA Cosmos 23:28 CAD to NVIDIA Isaac Sim Workflow 27:34 Isaac Assist for Isaac Sim Automation 31:10 Isaac Teleop and Apple Vision Pro Data Collection 34:43 NVIDIA Isaac Lab and GR00T Robot Training 38:28 Isaac ROS Visual SLAM Navigation 41:45 Newton Physics for Robot Simulation 52:32 Ada Roadmap and Healthcare Robotics 1:00:07 Robot Scoring, Feedback, and Overfitting 1:08:08 Synthetic Data for Healthcare Robotics 1:17:45 Human Control and Robot Safety 1:22:03 Physical AI Careers and Community 1:36:05 Developer Takeaways for Physical AI ---------------------------------------------------------------------------- List of Resources - 10 Things - https://10things.tech/ - Kimate Richards on LinkedIn - https://www.linkedin.com/in/kimaterichards - Oliver Choy on LinkedIn - https://www.linkedin.com/in/oliverchoy/ - Isaac Assist GitHub - https://github.com/k3street/Omniverse_Nemotron_Ext - Newton Physics - https://developer.nvidia.com/newton-physics - Isaac ROS Visual SLAM GitHub - https://github.com/NVIDIA-ISAAC-ROS/isaac_ros_visual_slam - NVIDIA Brev - https://brev.nvidia.com/ - SIGGRAPH Physical AI Playlist - https://www.youtube.com/playlist?list=PLViKEFsN3lPQ Continue the discussion and ask follow-up questions in the NVIDIA Omniverse Discord thread: https://discord.com/channels/827959428476174346/1539333016575549470 ---------------------------------------------------------------------------- 📆 Check out the full calendar for all of our upcoming events → https://nvda.ws/3JqaWnA ---------------------------------------------------------------------------- ⬇️Get Started → https://nvda.ws/4cZAZO1 👀Explore OpenUSD → https://nvda.ws/3CeozBQ 👥Join the Community → https://nvda.ws/3ZMfc6e #NVIDIA #NVIDIAOmniverse #NVIDIAIsaac #IsaacSim #IsaacROS #IsaacLab #NVIDIACosmos #Robotics #PhysicalAI #RobotSimulation #DigitalTwins #GR00T #OpenUSD #SyntheticData #DeveloperTools