Texture measurement and identification of object surface by MEMS tactile sensor

The surface texture of various materials has been characterized by tactile sensor with a micro-cantilever embedded in the PDMS elastomer. The maximum output (resistance of strain gauge on the cantilever) change of the sensor by indentation of the object on the sensor surface depends on the hardness or thickness of the object. Moreover, output change is asymmetric in a back-and-forth indentation test because of relaxation of deformation. On the other hand, the output increases rapidly at beginning of stepwise sliding of the sensor at the object surface because of static friction, and changes periodically corresponding to stick-slip oscillation or surface roughness after slipping. It is demonstrated that about 30 materials of papers, clothes, leathers, and plastic sheets, have been classified into 5 clusters by the principal component analysis using feature quantities extracted from the sensor outputs.