Studies of a Material Recognition System with an Artificial Finger Using a Pneumatic Cylinder

Abstract

We have developed an artificial finger which has a human tactile sensibility function, and is desired for some kinds of robots. This finger can recognizes some materials by only the actions of pressing and tracing an object actuated by the pneumatic cylinder. In general, there are many factors in recognizing the material, that is, “Surface condition”, “Hardness” and “Temperature” and they are used as input factors in the neural network in the study. This study also borrows from the action potential of humans. The action potential fires some impulses when a stimulus is applied to the skin. The human brain handles the impulse as input data. The advantage of inducing the impulse model in the artificial finger is to reduce the volume of calculation, and to handle the data of some sensors in parallel. Furthermore, we provided an artificial finger that has a finger joint. The pneumatic cylinder and the tendon actuate the joint to bend the finger. The purpose of making the finger joint is to improve the sensitivity for parameters and the effects are certified in this report.