NVIDIA/IsaacTeleop

What it solves

It addresses the data bottleneck in robot learning by providing a unified framework for collecting high-fidelity human demonstration data. It streamlines the process of integrating different devices and ensures that data is interoperable across different hardware platforms.

How it works

Isaac Teleop provides a standardized device interface and a flexible retargeting framework that works across both simulated and real-world teleoperation. It allows users to capture human movements—via XR headsets, gloves, or full-body tracking—and map those movements to robot grippers, tri-finger hands, or full-body loco-manipulation systems.

Who it’s for

It is designed for robotics researchers and developers who need to collect high-quality egocentric and robot demonstration data for training robot learning models.

Highlights

• Unified stack for both simulation and real-world teleoperation.
• Standardized device interface to simplify hardware integration.
• Support for various manipulation tasks, including gripper, tri-finger, and dex-hand manipulation.
• Capability for egocentric "no-robot" data collection.
• Integration with NVIDIA Isaac Lab.