Prognostic Methods for Predictive Maintenance: A generalized Topology

Prognostic methods for predictive maintenance have been presented extensively in the literature. While this area's continuing effort positively affects individual predictive maintenance solutions' performance and capabilities, a method's setup remains a big hurdle as the solution space is becoming more complex. The critical settings of a prognostic method are the selection of suitable modeling techniques used for behavior- and condition-modeling, as well as a forecast model for failure prediction. This paper presents a generalized topology of a prognostic method to ease the design of maintenance systems and allow for quicker individual method design and modification. After a broad literature review, the topology and its base components are presented, and an overview of the different kinds of models related to predictive maintenance applications is given. Copyright (C) 2021 The Authors.