A Novel Tactile Sensor for Slip Detection Based on Relative Angular Velocity

Abstract

Object slippage should be prevented as one of the major challenges in robotic manipulation. Applying insufficient grasp force leads to the object slip, whereas a large predetermined grasp force can damage sensitive and fragile objects. The ability to detect slippage in real-time implementations provides required feedback for grasp force regulation, which allows the hand (gripper) to reach the appropriate grasp force and prevent object slip. Predominantly during the slip sensation, high-frequency vibrations are generated due to the relative motion between the object and the fingers. In the present work, a novel tactile slip detection sensor was developed based on the effects of skin vibrations resulting from the slip and by measuring the relative angular velocity between the skin and the rigid parts of the finger. Substantial experimental studies revealed the proposed method can detect slippage events with a 100% success rate and low latency, which reveals merits of the proposed method and the fabricated sensor. Also, the sensor arrangement is designed with an emphasis on minimizing the noise resulting from the actuation motions and vibrations, and increasing the sensitivity to the slippage events.