Applying 3D Vision Sensing to an Industrial Robot for Vehicle Assembly under Noisy Environment

This paper presents a method for automated wheel hub assembly which is important work in the industrial applications. A key requirements for this work are the positioning accuracy of manipulator in working space and the 6D object's pose determined by the 3D vision algorithms. A noncontact method using 3D scanner mounted on the end-effector robot is utilized for providing measurements of object pose in workspace. By controlling the end-effector of robot moving to the working space, the errors of the end-effector are compensated directly. Some experiments to pick up the wheel hub using an industrial robot were implemented to validate the assembly performance of the proposed method.