Dynamic tracking of moving objects in microassembly through visual servoing

Abstract

Precise micropart alignment is a crucial factor in most gripper-based microassembly processes. For the micropart to be grasped or manipulated, these processes require the micropart to be positioned and oriented properly for the microgripper. At present, many of the these processes still rely on operators to monitor and align the micropart manually through visual images provided by the camera on top of the assembly line. However, due to the limited field of view of the microassembly system microscope, the micropart may move outside of the visual monitoring area at some point during the manipulation process. The present work proposes an integrated microassembly algorithm that performs the assembly process regardless of the micropart initial orientation. The algorithm automatically aligns and tracks the micropart during the manipulation process. As the micropart rotates to the required grasping orientation, the algorithm projects the future motion of the micropart, repositions it, and simultaneously, using a PID control algorithm, maintains the micropart within the field of view of the microscope. The proposed algorithm eliminates the need for a manual alignment process, which is time consuming and is subject to error. The algorithm was implemented and evaluated on an in-house 6 DOF microassembly manipulator. Experimental results confirmed that the proposed algorithm successfully tracked and corrected a 45-degree misaligned micropart at a specified location within the camera field of view with a steady-state error of +/−15 pixels.