A piezoelectret-based approach for touching and slipping detection in robotic hands

Abstract

Robotic hands are very essential to upper-limb amputees for their hand function restoration. However, most of the commercially available robotic hands do not have a sensation of touching or slipping, and thus it is not easy to properly grasp objects only with a visual feedback. Different sensing technologies have been proposed for tactile and slip signal detection in robotic hands, but practical applications are still limited because of either complicated realization or high cost. Piezoelectrets are a kind of novel transducer material, which would be a promising candidate for sensation signal acquisition in robotic hands partly due to their excellent piezoelectric response and polymer-based properties. In this work, a piezoelectret-based sensor unit was designed to detect the touching and slipping of a robotic hand, and the relevant sensing performance was quantitatively evaluated by experiments. The experimental results demonstrated that the piezoelectret-based sensor is a suitable approach for both touching and slipping detection in robotic hands, and a preliminary involuntary control of robotic hands may be realized with an appropriately designed control strategy.