Towards comprehensive capture of human grasping and manipulation skills

Grasping plays a central role in our daily life. To interact with objects surrounding them, people use a large diversity of hand configurations in combination with forces ranging from the small ones involved in manipulating a pen for writing, to larger forces such as when drinking a cup full of water, and even larger ones such as when wielding a hammer. In this paper we present a setup to capture human hand configuration and motion as well as the forces applied by the hand on objects while performing a task. Hand configuration is obtained through the use of a data glove device while interaction forces are measured through an array of tactile sensors. Current approaches in the state-of-the-art are limited in that they only measure interaction forces on the fingers or the palm, ignoring the important role of the sides of the fingers in achieving a grasp/manipulation task. We propose a new setup for a “sensorized” data glove to address these limitations and through which a more complete picture of human hand response in grasping and manipulation can be obtained. This setup was successfully tested on five subjects performing a variety of different tasks.