Visual-servoing of a six-degree-of-freedom robotic manipulator for automated microassembly task execution

Abstract

This paper introduces a visual-servoing system used in the automation of microelectromechanical system microassembly. In the proposed system, small microparts, 20 µm wide and 2 µm thick, are automatically grasped using a passive compliant microgripper in a first step toward achieving complete automation of the microassembly process. A fuzzy logic controller uses vision-based position and force feedback to correct the microgripper and micropart alignment and to guide the system toward achievement of a successful grasp. A depth-from-defocus depth estimation technique is used to properly align the micropart in the same horizontal plane as the microgripper. The performance of the system is investigated experimentally, and experimental results are presented.