Contact Identification using Tactile Arrays

The proposed method allows a robot to quantify tactile recognition of grasped objects. The responses by the robot are matched by randomly introducing objects following the completion of tactile sensor based training phases on 4 specific objects. The reproduction of the robotic contact identification is accomplished by matching threshold values of the smallest eigenvalue of the Quadric surface - called the eigenvalue trajectory. Once the grasped objects are introduced to the 2-fingered robot hand, the eigenvalue trajectory of contact surfaces are calculated by the system. Calculations are based on the degree of similarity of the characteristics between the newly received and previously trained tactile data. Experimental results indicate that the proposed algorithm allows a successful quantification of robotic contact recognition and the accomplishment of high levels of accuracy in contact identification. These results confirm that the proposed method is useful for identifying contact surfaces based on tactile sensor data.