| Article ID | Journal | Published Year | Pages | File Type |
|---|---|---|---|---|
| 413415 | Robotics and Autonomous Systems | 2014 | 7 Pages |
In dexterous robotic manipulation, it is essential to control the force exerted by the robot hands while grasping. This paper describes a method by which robot hands can be controlled on the basis of previous experience of slippage of objects held by the hand. We developed an anthropomorphic human scale robot hand equipped with an elastic skin in which two types of sensor are randomly embedded. One of these sensors is a piezoelectric polyvinylidenefluoride (PVDF) film which can be used for the detection of pressure changes. The other is a strain gauge which can measure static pressure. In our system, PVDF films are used to detect slipping, and strain gauges to measure stresses which are caused by normal and shear forces. The stress measured by the strain gauges is used as input data to a neural network which controls the actuators of the robot. Once slip is detected, the neural network is updated to prevent it. We show that this system can control the grasp force of the robot hand and adapt it to the weight of the object. By using this method, it was shown that robots can hold objects safely.
► We propose a method to predict slips which occur between a robot hand and an object on the basis of previous experience. ► We use elastic artificial skin in which two types of sensor are randomly embedded. ► PVDF films are used to detect slips and strain gauges are used to predict slips. ► The neural network to control the grasp force is updated once a slip is detected. ► We show that the network was trained to control the grasp force appropriately.
