Automation and increased repeatability for quality verification measurements on espresso coffee.

In the dynamic context of the modern coffee industry, the integration of Industry 4.0 methodologies has emerged as an indispensable complement to traditional manual labor, fostering automation and heightened precision. This thesis, titled "Increase of repeatability and automation in coffee quality measurements," explores the successful automation of critical processes involved in the assessment of key coffee quality parameters. A prototype, utilizing a vertical automatic guide steered by a stepper motor, was developed to facilitate the measurement of temperature, resulting in improved measurement repeatability. The subsequent stages of the research delve into the augmentation of automation in additional parameters, focusing on persistency, crema volume, and crema quality. Employing digital image processing techniques, the thesis attains enhanced precision and reduced execution times for crema volume measurements, a pivotal element in the evaluation of coffee quality. Moreover, a robust artificial intelligence model, specifically a deep neural network employing ternary classification, was developed to distinguish between superior, average, and substandard crema, thereby significantly contributing to the objective of ensuring consistent and high-quality coffee production. This multifaceted approach not only accentuates the precision and repeatability of measurements but also effectively curtails coffee wastage by executing sequential measurements on a single sample, in contrast to the current approach of conducting multiple measurements on diverse samples for each assessment parameter.