Abstract: An apparatus, method, and non-transitory machine-readable medium provide for improved feature extraction and fault detection in a non-stationary process through unsupervised machine learning. The apparatus includes a memory and a processor operably connected to the memory. The processor receives training data regarding a field device in an industrial process control and automation system; extracts a meaningful feature from the training data; performs an unsupervised classification to determine a health index for the meaningful feature; identifies a faulty condition of real-time data using the health index of the meaningful feature; and performs a rectifying operation in the industrial process control and automation system for correcting the faulty condition of the field device.

Abstract: An apparatus, method, and non-transitory machine-readable medium provide for improved feature extraction and fault detection in a non-stationary process through unsupervised machine learning. The apparatus includes a memory and a processor operably connected to the memory. The processor receives training data regarding a field device in an industrial process control and automation system; extracts a meaningful feature from the training data; performs an unsupervised classification to determine a health index for the meaningful feature; identifies a faulty condition of real-time data using the health index of the meaningful feature; and performs a rectifying operation in the industrial process control and automation system for correcting the faulty condition of the field device.

Abstract: A framework for autonomous predictive health monitoring includes online monitoring, offline training, and self-learning components. The monitoring component includes analyzing streaming incoming process data, which includes process variable and key performance indicators (KPIs), from multiple sources, in real time, to determine an overall health index, determine faults, diagnose and isolate faulty process variables that contribute to the health index, and predict a trend and a magnitude of the health index before failure. The self-learning component includes services linked to event management, to correct the health index from probabilities calculated based on operator feedback on true or false events after analyzing each of the detected events, self-tune limits and other model parameters, and trigger training of a model when a new normal pattern is detected.