X Square Robot's TwinDEX Wearable Collector Boosts Dexterous Data Throughput 5.3x
FellMentKE · x · 2026-09-02
Data collection is a major bottleneck in dexterous robot learning: on-robot teleoperation yields high-quality demos but is expensive, slow to reset, and requires the robot every time. Chinese startup X Square Robot's TwinDEX instead uses a wearable robotic end-effector to collect demonstrations from humans, then deploys the learned policy on a matched robot hand. Crucially it's not just motion tracking — the wearable and hand are co-designed to match kinematics, contact geometry, appearance, sensing, and timing. In evaluation, TwinDEX delivered up to 5.3x the effective throughput of on-robot teleoperation — an interesting route to scaling dexterous data.
More from Embodied
- SimInfinity: open-source robotic simulation benchmark with 10K real-world tasks — joemeno · 2026-10-18
- Robi: a German/English voice companion demo built on Gemma, Tyger and Humaneness Voice TTS — realmrfakename · 2026-10-18
- Munich's Plim unveils wheeled home-care humanoid controllable from 9,000 km away — CyberRobooo · 2026-10-18
- Musk says Optimus and Cybercab will make up the vast majority of Tesla's value — XFreeze · 2026-10-18
- SAE weight editing steers robot behavior at inference time, IIIT Hyderabad paper lands IJCAI Safe Physical AI workshop — ponguru · 2026-10-18
- NTU-led survey reviews nearly 200 papers on affordance learning for embodied AI — jiqizhixin · 2026-10-18