Data Glove

A wearable sensing glove that tracks each finger's bend and the hand's pose in real time.

A data glove is a wearable input device that measures each finger joint's angle using bend sensors, an inertial measurement unit (IMU — a chip that senses angular velocity and acceleration), or electromagnetic sensing, usually paired with a tracker to also capture the whole hand's pose in space. It was originally used mainly for human-computer interaction, virtual reality, and animation; the 1977 Sayre Glove is considered the first one. In embodied AI it has two main uses: teleoperating a dexterous hand by mapping a human hand's motion onto the robot hand in real time, and collecting human hand-manipulation data directly, without any robot present, which is then converted through motion retargeting — translating human motion into robot joint commands — for training. Versions with force feedback can also relay the sensation of contact back to the operator.

ExampleMANUS markets its Metagloves Pro for robot-training data collection and teleoperation. The DexCap system pairs an electromagnetic motion-capture glove with a SLAM camera — one that can compute its own position while moving — to track the wrist, giving a portable way to record human manipulation for dexterous hands to learn from through imitation learning.

Also called

Wired Glove, CyberGlove

Related

Motion Capture · Haptic Glove · Tactile Glove · Teleoperation · Motion Retargeting · Dexterous Hand

Sources

Wikipedia: Wired glove
MANUS 官网（Metagloves 产品） (Chinese)
DexCap: Scalable and Portable Mocap Data Collection System for Dexterous Manipulation (arXiv 2403.07788)

As of

2026-09