Fuzzy Logic Based Sensor Skin for Robotic Applications

摘要

Tactile sensors play an important role in many robotic applications providing information about shape, roughness, frictions and acting forces. However, it is challenging cover large robots with tactile sensors. In this paper we introduce a modular tactile sensor skin built on master and slave principle. The structure of the sensor skin was inspired by the biological multi layered skin from humans. Modular design principle allows covering different flat and rough robot geometries. The skin is able to detect tactile and pressure sensor values and is therefore applicable to different applications. Fuzzy logic based data fusion method allows the input values to be fuzzy while receiving numerical values as the output. This concept reduces the number of required sensor to a minimum and consequently also the computational effort and production costs.