Abstract
In this paper, a robot identification system for different materials using polyvinylidene fluoride (PVDF) soft sensor is developed. The sensor, imitating the finger-pads, was constructed from brass traveler as an effective substitution fingerprint, a silicon rubber (SYLGARD 184) sheet as elastic skin and PVDF film as the sensing element. It is designed for high imitation of human natural and high reliability. PVDF film is chosen for the flexibility and corrosion resistance. Furthermore, the output signal of the sensor is obtained by contacting and scanning objects. Since the signals from sensor are complex, the Robot Signal Process System (RSPS) is introduced to differentiate materials and classify the objects both qualitatively and quantitatively. Several experiments using the transformed robot are presented in the study. The results show that the robot system is effective (95%) and reliable in discriminating objects depending on material properties such as textures.