Abstract:
This paper presents modifications made to the Adaptive Capacitive Sensor for Obstacle Ranging (ACSOR) [1] that have enabled the sensor to provide material type data for an object in the sensing field. The ACSOR, previously capable of obstacle ranging (up to 500mm) in various densities of particle laden air, has been modified and fitted with a noise and stimuliresponse analyzing algorithm allowing the sensor to determine the material type of a sensed object. Experimental results have demonstrated the modified sensor's ability to successfully classify the material type of several different objects whilst fitted to an inmotion 6-DOF anthropomorphic robotic arm in typical indoors conditions; 3.56% incorrect classifications (from a set of known material types) with an unintelligent 'hard logic' classifier.
