Estudo do Método de Inspeção Visual e Automática do Número de Identificação Veicular de Chassis de Motocicletas

Abstract

The inspection method employed in validating the chassis number after it has been recorded does not always meet the quality requirements required by the market. In the evaluated process, the inspection is performed manually, where an operator makes the visual conference, submitting to a large automotive industry plant located in the Manaus Industrial Pole (PIM) market leader in the two-wheel market the risk of selling a motorcycle with registration different from the documentation. Thus, this work develops an automatic inspection system, applied to the quality control of the vehicle identification number of motorcycles. And with this it was possible to evaluate its reliability of the data generated in order to provide resolution to the recording problems. Given the implementation of the development process of this project it was possible to clearly observe and identify problems resulting from the printing of the motorcycle VIN engraving. Regarding the detection of the proposed method with the human inspection method, it can be said that the results achieved are superior, evidencing the robustness of the automatic inspection system, since it was possible to identify very accurately all the defects. In addition, hit rates are higher than those obtained by human operators. Thus, the obtained results allowed a significant improvement in the total production time of the model, as it was possible to reduce 80% of the time spent to perform the inspection activity and to reduce the internal defect rates.The inspection method employed in validating the chassis number after it has been recorded does not always meet the quality requirements required by the market. In the evaluated process, the inspection is performed manually, where an operator makes the visual conference, submitting to a large automotive industry plant located in the Manaus Industrial Pole (PIM) market leader in the two-wheel market the risk of selling a motorcycle with registration different from the documentation. Thus, this work develops an automatic inspection system, applied to the quality control of the vehicle identification number of motorcycles. And with this it was possible to evaluate its reliability of the data generated in order to provide resolution to the recording problems. Given the implementation of the development process of this project it was possible to clearly observe and identify problems resulting from the printing of the motorcycle VIN engraving. Regarding the detection of the proposed method with the human inspection method, it can be said that the results achieved are superior, evidencing the robustness of the automatic inspection system, since it was possible to identify very accurately all the defects. In addition, hit rates are higher than those obtained by human operators. Thus, the obtained results allowed a significant improvement in the total production time of the model, as it was possible to reduce 80% of the time spent to perform the inspection activity and to reduce the internal defect rates.