Integration of artificial intelligence towards increased degree of utilization of wearable robotic gloves

Abstract

Recently, soft wearable robotic gloves have been widely developed to assist people with a loss of hand mobility. In this study, we propose several learning-based algorithms towards increasing the degree of utilization for these soft wearable robotic gloves. In particular, the study focuses on two major areas of utilization for these devices: grasping objects of various physical properties found in daily life and utilizing the soft robotic gloves for hand rehabilitation. Therefore, a method to predict fingertip forces applied to deformable objects is proposed and applied to a soft wearable robotic glove. Then, to enable stroke patients to actively practice various hand postures related to activities of daily living, a vision-based framework for detecting users’ intentions’ regarding multiple hand postures is proposed.