Abstract
In this letter, we present a data-driven condition-monitoring system for a moka pot aiming at anomaly detection in the coffee-preparation process. A data-acquisition system and the corresponding generation process of a comprehensive dataset (including data from ideal and anomalous brewing scenarios) are described. Supervised and unsupervised machine learning algorithms are trained and tested on the dataset aiming at detecting anomalies in the process and showing the relevance of the considered framework.