Degradation modeling for real-time estimation of residual lifetimes in dynamic environments

Abstract

This article presents a methodology for modeling degradation signals from components functioning under dynamically evolving environment conditions. In situ sensor signals related to the degradation process are utilized as well as the environment conditions, to predict and update, in real-time, the distribution of a component’s residual lifetime. The model assumes that the time-dependent rate at which a component’s degradation signal increases (or decreases) is affected by the severity of the current environmental or operational conditions. These conditions are assumed to evolve as a continuous-time Markov chain. Unique to the proposed model is the union of historical data with real-time, sensor-based data to update the signal parameters, environment parameters, and the residual lifetime distribution of the component within a Bayesian framework.

Keywords:

• Degradation model
• residual lifetime distribution
• reliability
• Bayesian estimation

Additional information

Notes on contributors

Linkan Bian

Linkan Bian received his Ph.D. in Industrial and Systems Engineering from Georgia Institute of Technology in 2013. He received his dual M.S. degree in Statistics and Mathematics from Michigan State University in 2008. He also received his B.S. degree in Applied Mathematics from Beijing University in 2005. He joined the Industrial and Systems Engineering Department as an Assistant Professor in August 2013. His main areas of research interest are in system reliability and sustainability, diagnosis and prognostics, stochastic models, and statistical learning, with applications in automotive systems, wind power systems, advanced manufacturing systems, sensor networks, and other complex engineering systems. Dr. Bian’s publications have appeared in journals such as IIE Transactions, Statistical Analysis and Data Mining, Naval Research Logistics, and several conference proceedings. He is a member of the Institute for Operations Research and the Management Sciences (INFORMS) and the Institute of Industrial Engineers (IIE).

Nagi Gebraeel

Nagi Gebraeel is a Chandler Family Associate Professor in the Stewart School of Industrial & Systems Engineering. He received his M.S. and Ph.D. from Purdue University in 1998 and 2003, respectively. Dr. Gebraeel’s research focuses on prognostics and predictive analytics for improving reliability and sustainability by leveraging sensor data streams. He is also interested in integrating predictive analytics with operational and logistical decision-making strategies in the manufacturing and service sectors. Dr. Gebraeel is a member of the Institute of Industrial Engineers (IIE), Institute for Operations Research and the Management Sciences (INFORMS), and The American Institute of Aeronautics and Astronautics (AIAA).

Jeffrey P. Kharoufeh

Jeffrey P. Kharoufeh is an Associate Professor in the Department of Industrial Engineering at the University of Pittsburgh. He specializes in the application of probability and stochastic processes to the design, performance evaluation, control, and optimization of stochastic engineering and service systems. His focus areas include queueing systems, reliability theory, maintenance optimization, and models for computer and communications systems. He earned B.S. and M.S. degrees in Industrial & Systems Engineering from Ohio University and a Ph.D. in Industrial Engineering & Operations Research from the Pennsylvania State University. Professor Kharoufeh currently serves as Area Editor for Operations Research Letters and the Wiley Encyclopedia of Operations Research and Management Science and as Associate Editor for Operations Research. He is a Senior Member of the Institute of Industrial Engineers (IIE) and a professional member of INFORMS and the Applied Probability Society.