Diagsence is aimed at transforming sensory output during normal mode and learning operational persistence through correlation analysis of sensory data to detect anomalies. In contrast with the engineering systems based on huge databases of incidences, the technology adopted by Diagsence separates the norms statistically so that an alarm is only triggered when there is a real departure from the expected behavior, maintaining all operations efficiently and without the need for constant calibration.