Data Gloves
High-DOF hand motion and tactile capture — build rich dexterous manipulation datasets.
Overview
The data gloves capture high-degree-of-freedom hand motion through distributed bend sensors, together with fingertip tactile feedback. Low-latency wireless streaming stays hardware-synchronized with the camera feeds, so every finger movement and contact force is aligned frame-by-frame for dexterous-manipulation learning.
Key Features
15+ degrees of freedom of finger joint angles per hand.
Fingertip tactile and pressure sensing at every fingertip.
Low-latency wireless streaming, hardware-synced with camera feeds.
Flexible, breathable material that fits a wide range of hand sizes.
Specifications
- DOF
- 15+ per hand
- Sample rate
- 120 Hz
- Tactile points
- 5 fingertips
- Latency
- <10 ms
- Connectivity
- 2.4G wireless / BLE
- Battery
- ~6 hours
Use Cases
Teleoperation of dexterous robotic hands.
Imitation learning of hand motion.
Multi-modal tactile-visual data collection.