Aplicação de inferência fuzzy para tomada de decisão em processos de SMT: um caso no PIM

Abstract

Surface Mount Technology (SMT) is widely used in the production of electronic components, especially in the Manaus Industrial Pole (PIM), one of the largest electronics production centers in Brazil. However, its complexity and sensitivity to parameter variations, such as temperature and alignment, pose significant challenges to process accuracy and consistency. Traditional control methods, such as PID, often fail to meet the adaptability and efficiency demands required in these scenarios. Fuzzy logic emerges as a promising alternative, providing support in situations of uncertainty and variability. This study developed and implemented a fuzzy inference system for controlling critical parameters in SMT processes at PIM. The methodology involved identifying key variables, such as temperature and pressure, and constructing a fuzzy model based on linguistic rules and expert knowledge. The system was validated through comparative analysis before and after its implementation, evaluating aspects such as production efficiency, rework rate, and final product quality. The application of the fuzzy system resulted in significant improvements in the efficiency of the SMT process. There was a reduction in assembly failures, less need for rework, and greater consistency in production. In addition, the system demonstrated greater adaptability to operational variations compared to traditional control methods, contributing to more sustainable and efficient production. The use of fuzzy inference systems in SMT processes at PIM proved to be an effective approach for optimizing decision-making, improving quality and production efficiency. This research reinforces the role of fuzzy logic as a viable tool for meeting the demands of Industry 4.0, promoting innovations in industrial automation, and contributing to the competitiveness of PIM companies.