Stanford's DITTO-X does forward and reverse teleoperation for dexterous robot hands
chris_j_paxton · x · 2026-10-06
Stanford researchers present DITTO-X, an actuated exoskeleton interface for dexterous teleoperation that works across Sharpa Wave, Wuji 2, and Inspire hands without per-hand redesign.
- Forward teleoperation with force and haptic feedback yields higher-quality demonstrations for contact-rich manipulation
- Reverse teleoperation lets the robot drive the operator's fingers to match its own state, enabling fluent mid-task human takeover without unstable grasps
- These interventions become DAgger data that improves deployed policies
- Diffusion policies trained on DITTO-X demos complete six tasks (battery insertion, cube in drawer, pick-and-place, etc.) in real time
- Paper, code, and dataset released
More from Embodied
- Black Forest Labs open-sources 7B FLUX 3 Action, tops RoboLab-120 at 3.95x speed — dl_weekly · 2026-10-06
- Muse Gadgets Open-Sources ESP32 Firmware and Linux SDK for Building Muse-Compatible Hardware — qinzytech · 2026-10-06
- RAI Institute's new bipedal Roadrunner robot debuts at IROS, built from idea to prototype in 6 months — shekitup · 2026-10-06
- NVIDIA VP Rev Lebaredian: digital twins let robots learn physics in simulation — RevLebaredian · 2026-10-06
- NormalFactory unveils FrontierCAD, a verifier for grading AI-generated CAD parts and assemblies — hudzah · 2026-10-06
- Walden Robotics CTO to discuss Physical AI for line troubleshooting at Operations Calling — adnothing · 2026-10-06