AI enabled Computer Vision Cobot for Real-time Screw position detector and Screw tightener used for Smart Factory

Abstract

Almost in every assembly line in the most of the machine workshop has a common part, i.e., a large number of screws to be fitted. Moreover, the quality of the product is mostly dependent on how good fitted the screws in that. However, in most of the assembling plant, screws are tightening manually in the products like gear box assembling, breaking system assembling etc. Sometimes this becomes one of the major reasons behind a product breakdown. A manual fitted product can never compete with a machine fitted product. Developing Cobot for screwing is indeed a great challenge, since the detection of screw hole is not a simple task for an unaligned item without an intelligent system. In this work, we have developed an intelligence Cobot integrated with Camera. Using Machine Learning algorithm the camera module can detect the hole’s position accurately in real-time and export the coordinate data of holes position to a SCADA Cobot to places the appropriate screw and to tighten the screw with pre-set exact amount of torque. Therefore, the quality of the product will get highly improved in comparison to manual assembling.