Data Driven Methods for the Prediction of Failures

Abstract

The reliability and safety operation of an industrial system are the main objectives of industrial companies to remain competitive in a constantly growing market. Unexpected shutdowns can often lead to physical hazards as well as economic consequences in key sectors. Hence, fault prediction emerges as an important focus of the industry. Thus, this paper aims to detail the prognostic aspect and provides a state of the art of existing data-driven prognostic methods used in the literature. This paper shows the diversity of possible prognostic methods and the choice of one among them that will define a framework for industrials.