# Isaac Capture > Isaac Capture is shared capture infrastructure for teleoperation (Isaac Teleop) and egocentric data (Ego Pipeline): device plugins, retargeters, and data utilities for recording, replaying, and converting human and robot demonstration data. 2025-2026, NVIDIA CORPORATION & AFFILIATES, last updated on November 18, 2026 ## Pages in this subsection - [Isaac Lab & Sim](isaac_lab_sim.md): This integration guide is being prepared. - [Isaac ROS](isaac_ros.md): Isaac ROS Teleop consumes `teleop_ros2`, the reference ROS 2 publisher for Isaac Teleop - [LeRobot and SO-101](lerobot/index.md): With Isaac Teleop, we got an end-to-end data collection and training pipeline for SO-101 in both sim... - [Supported Teleop devices](lerobot/devices.md): Isaac Teleop drives the SO-101 from two input devices today. Both command the same SO-101 follower - [Data Collection in Real](lerobot/data_collection_real.md): Record demonstrations on a physical SO-101 into a LeRobot dataset, driving the follower with either ... - [Data Collection in Sim](lerobot/data_collection_sim.md): Collect SO-101 demonstrations in simulation with NVIDIA Isaac Lab, on the cube-stacking task. You dr... - [Model Training with GR00T](lerobot/training_groot.md): With a LeRobot dataset collected on real hardware (Data Collection in Real) or in simulation - [reBot B601-RS on Jetson AGX Orin](lerobot/rebot.md): The reBot B601-RS is a 6-DOF arm built from ## Optional - [Top-level llms.txt](../llms.txt): Complete documentation index.