A Digital Failure Twin Model For PHM: From Concepts To Maturity Levels

In the context of industry 4.0, traditional reliability engineering has the opportunity to greatly benefit from the big data generated by Industrial Internet of Things (IIoT) and the advancement of industrial intelligence technologies. To make full use of the opportunities emerging in the era of industry 4.0, a Digital Failure Twin (DFT) is proposed as a dynamic virtual modeling framework that encompasses various dimensions such as model, data, connection, and computation. This framework is designed to simulate and forecast the failure behavior of complex real-world systems throughout their entire life cycle, enhancing Prognostics and Health Management (PHM) application, ultimately leading to improved reliability, safety, and operational efficiency. Five maturity levels of a DFT model are defined to support the development and implementation of DFT in practice. This evolution entails transitioning from static failure modeling based on historical data to dynamic predictive maintenance capabilities integrated with artificial intelligence. The potential application of the DFT framework on prognostics and health management is also discussed. Through the discussion, we show that introducing the DFT in prognostics and health management could improve operational and maintenance planning and reduce failure occurrence, as the framework could provide an accurate and up-to-date prediction of the evolution of the failure behavior.