Origami Robotics launches S1 glove for robot-learning data

Origami Robotics has introduced S1, a wearable system for capturing the details of human hand movements for robot training.

What happened

S1 records 20 finger-joint angles, pressure at the fingertips and palm, wrist movement and synchronized video. The glove was designed alongside Origami's robot hand so that captured finger positions correspond closely to joints the robot can command.

Production of the glove has started, although transferring complex human dexterity into reliable robot execution remains a work in progress.

Why it matters

Camera-only demonstrations lose information when fingers are hidden behind objects. They also cannot reliably measure how firmly someone is gripping or pressing.

Instrumented gloves can preserve motion and contact information that would otherwise be missing from a robot-learning dataset.

The bigger picture

Robotics teams are increasingly designing data-collection hardware together with the robots that will use the data. This could reduce the difficult translation step between a human demonstration and a robot's physical body.