Abstract
The purpose of the present research is to build a tactile sensing system for robots inspired by biological cognitive systems. In information processing of a biological organism, there is stochastic resonance capable of enhancing sensitivity by making a proper noise superimpose upon not detectable weak signal to detect the signal. Meanwhile, since in tactile sensing a noise gets mixed inevitably by contact with an object and sensor's movement on it, it is expected that sensing accuracy can be increased by mimicking the biological information processing. In the present research, we develop an object surface scanning system consisting of a tactile sensor, a liner motor, and an automatic stage. The system is utilized in a series of experiments to verify the validity of stochastic resonance in tactile sensing. It is found that a subtle concave and convex portions on the object surface are measured by the stochastic resonance.
