Predictive Analytic Framework for Machine State Classification and Remaining Useful Life Prediction Using Machine Learning

Abstract

Predictive analytics has found a predominant place in almost all the fields. It has transformed the perspective of the industrial maintenance activities from preventive to predictive maintenance. Predictive maintenance focus on scheduling the maintenance activity before the onset of failures to reduce the downtime of equipment and to mitigate the effect of failures. This accelerated the deployment of Artificial Intelligence and Machine Learning based prognostic models, to predict the faults in the equipment under study. The existing models are backed by intense feature engineering, which inherently has the threat of overlooking the important features of the data newline