Remote quality control of welded joints of films by measuring defects by artificial intelligence system

A.S. Stepashkina, К.V. Epifantsev,V.O. Smirnova, O.V. Chuprinova, K.V. Zolotukhin

St. Petersburg State University of Aerospace Instrumentation, St. Petersburg, Russia,

Al’manac of Modern Metrology № 4 (32) 2022, pages 75–88

Abstract. The issues of non-contact measurements are important at the stage of quality control in order to speed up the process of detecting a defect and deter­mining its accuracy parameters. The contactless (remote) method of quality con­trol is also applicable for sterile types of products, direct contact of the instrument with which is extremely undesirable. Modern systems of optical measuring mech­anisms, such as video measuring machines, are certainly capable of detecting a defect with high accuracy from several samples presented. However, the low speed and the impossibility of intellectual sorting of defective elements from the corresponding qualitative characteristics limits their performance.
The paper describes the main methods of working with machine vision and suggests an approach to quality control of welded joint materials. Machine vision allows for continuous automated control in production. Machine vision is based on the processing and analysis of photographic materials, which serve as a basis for comparing products for defects. The introduction of such technology in pro­duction reduces the possible release of marriage by an order of magnitude. For implementation, the approach of the capillary control method and its implemen­tation using an algorithm in the KNIME analytical platform are considered. The capillary method is based on the use of a contrast agent, which, penetrating into various defects, forms visualized indicator patterns under the influence of capillary pressure. The control is carried out using machine vision: all materials are scanned in a continuous mode and compared with the created database of defects.
The described method will provide an improvement in the quality of the final product and will allow moving from manual defect checking to automated.

Keywords: automation, machine vision, smart manufacturing, product quality control, capillary method, robotic systems, welded joints.

Full texts of articles are available only in Russian in printed issues of the magazine.