Towards Developing an Algorithm in Discriminating Thin Material using Tri-axial Tactile Data from Human Tactile Sensation

In order to increase the performance and efficiency of current tactile sensors, the study on the versatility of the human hand and tactile function could provide excellent hint. Our objective is to develop a new algorithm based on comparing the behavior of human tactile mechanism (using psychophysics experiment) vs. a robotic hand (equipped with 3-axis tactile sensor) using stainless steel (SUS) foils and 1,000-yen bills as medium. The significance of tangential load is highlighted to distinguish between material thicknesses. The result from the first experiment shows that using normal vs. angled load, the Weber fraction is high during low thickness and decreased with increase of thickness. In the second experiment, we find that the ability to distinguish between sheet numbers improves by increasing tangential load compared to just normal load. Overall, it is noted that the tangential load makes the tactile sensor distinguish extremely thin material.