Uso de sensores a fibra óptica como ferramenta para predição de odor durante o processo de fermentação do café

Abstract

This study explores the use of fiber optic Mach-Zehnder interferometers (MZIs) to monitor the coffee bean fermentation process, with an emphasis on identifying variations in odor patterns. The analysis of CO2 and temperature data collected throughout the fermentation process, along with aroma profiles, revealed significant correlations that allow the detection of critical transition points between different aromas. The application of Principal Component Analysis (PCA) effectively reduced the complexity of the data, preserving a large portion of the original variability and enabling a clear differentiation between aromas. The results indicate that integrating data from optical and electronic sensors can optimize the monitoring of the fermentation process, offering an efficient and precise alternative to traditional quality control methods. The study suggests the need for future research to integrate additional sensors and more advanced machine learning algorithms to improve aroma classification and sensor repeatability.