Validação de componentes automobilísticos utilizando inteligência artificial

Abstract

As the quality standards of the automotive industry are constantly increasing, it is also necessary to increase the quality of the parts produced and shipped to the automobile manufacturers. For air conditioning pipes, it is of vital importance to install specific components that increase their useful life, and it is also very important to deliver these pipes consistently to the customer. It was therefore decided to improve the production line for air-conditioning pipes by developing an automated final inspection bench assisted by vision sensors. As soon as there is a defective part or a part without the necessary components, it is identified by the system and the error is corrected, thus nonconforming parts do not reach the customer. To find the optimal ope ration mode for the sensors, tests were developed with the color of the image capture base, the operation mode of the sensors, the use or not of the polarizing filter and the automatic position adjustment, then these tests were validated through weeks of continuous operation of the production line. As a result, it was observed that the designed system improved reliability, guaranteeing the destination of parts, in conformity and with all their components, to the customer.