Keynote Speakers

Dr. Jay Lee
Ohio Eminent Scholar, L.W. Scott Alter Chair, and Distinguished Univ. Professor Univ. of Cincinnati &
Director NSF Multi-Campus Industry/University Cooperative Research Center on Intelligent Maintenance Systems (IMS), Univ. of Cincinnati, Univ. of Michigan, Missouri Univ. of S&T and Univ. of Texas-Austin

Subject: ”Design of CPS-based Digital-Twin PHM in Future Industrial Systems”

Abstract
In today’s competitive business environment, companies are facing challenges in dealing with big data issues for rapid decision making for improved productivity. Many manufacturing systems are not ready to manage big data due to the lack of smart analytics tools. U.S. has been driving the Cyber Physical Systems (CPS), Industrial Internet, and Advanced Manufacturing Partnership (AMP) Program to advance future manufacturing. Germany is leading a transformation toward 4th Generation Industrial Revolution (Industry 4.0) based on Cyber-Physical Production System (CPPS). China has just launched 2025 Plan and Internet Plus to focus on strengthening manufacturing and accelerate service innovation. It is clear that as more predictive analytics software and embedded IoT are integrated in industrial products and systems, predictive technologies can further intertwine intelligent algorithms with electronics and tether-free intelligence to predict product performance degradation and autonomously manage and optimize product service needs.
The presentation will address the trends of predictive big data analytics and CPS for future industrial PHM systems. First, Cyber-Physical System (CPS) enabled PHM will be introduced. Second, advanced predictive analytics technologies for smart maintenance with case studies will be presented. Finally, a Digital-Twin PHM platform for a closed-loop product design will be discussed.


Prof. Jerome Antoni
Laboratoire Vibrations Acoustique, University of Lyon, France

Subject: "Diagnostic indicators in vibration-based Condition Monitoring: connections between cyclostationarity, higher-order statistics, and sparsity"

Abstract
Vibration-based Condition Monitoring of machines strongly relies on advanced signal processing tools. The currently available panel of cutting edge methods reflects several historical trends, driven by research in higher-order statistics, time-frequency and time-scale representations, cyclostationarity, and more recently sparsity, to cite a few. A common goal of these approaches is to extract the diagnostic information from the signals of interest in the form of a nonlinear or a nonstationary signature. The object of this presentation is twofold.
First, it intends to evidence new links between diagnostic criteria based on higher-order statistics (such as the kurtosis and its extensions like the spectral kurtosis), sparsity (on which the literature is rapidly growing) and cyclostationarity (such as the envelope spectrum). In particular, it is shown that the cyclostationary framework allows a very sparse representation of signals and contains the same information as delivered by the kurtosis and some measures of sparsity.
Next, based on these findings, a renewed interest is given to cyclostationarity with an effort to push back its current limitations. It is generalized to the case of machines operating under varying regime and a fast estimator of the Spectral Correlation Density – a key tool in cyclostationarity -- is also introduced, from which new diagnostic indicators are derived. Eventually, several examples of applications are presented in different areas of Condition Monitoring, including the automotive, aeronautic, and energy industries.

Prof. Steven Li
Department of Industrial Engineering and Engineering Management, Western New England University

Subject: "Issues and Challenges of Bayesian Inference in PHM: Prior, Data, or Lying"

Abstract
This talk is concerned with the issues and challenges of Bayesian statistical inference based decision making using disparate prior knowledge and extreme data. The effects of Bayesian inference due to rational and misspecified prior information as well as both extremely limited and abundant data are investigated. Taking system reliability analysis of complex engineering systems as examples, various Bayesian methods for system reliability analysis by effectively integrating various available sources of data and expert knowledge at both the subsystem and system levels are demonstrated. More specifically, three scenarios based on available information for the test data of a system and/or subsystems are studied using Bayesian inference techniques. The research also proposes the Bayesian melding method for integrating subsystem level priors with system level priors for both system and subsystem level reliability analysis. System reliability analysis results are compared between the Bayesian Melding method and the traditional approaches relying on system structure alone. Computational challenges for posterior inferences using the sophisticated Bayesian Melding method are addressed using Markov Chain Monte Carlo (MCMC) and adaptive Sampling Importance Re-sampling (SIR) methods. A variety of numerical examples with simulation results illustrate the applications of the proposed methods and provide insights for system reliability analysis using multilevel information.


Prof. Yaguo Lei
School of Mechanical Engineering, Xi’an Jiaotong University

Subject: "Machinery Health Monitoring and Intelligent Fault Diagnosis in Big Data Era"

Abstract
In order to fully inspect the health conditions of machinery, condition monitoring systems are used to collect real-time data from the machinery. Because the amount of the machines diagnosed is great and the number of the sensors for each machine is large, massive data are acquired by the high sampling frequency after the long-time monitoring. Such massive data make the machinery health monitoring endorse the concept and power of big data. To handle the mechanical big data, machinery intelligent fault diagnosis may be a promising tool since it is able to rapidly and efficiently process collected signals and provide accurate fault diagnosis results. This talk will first introduce the opportunities and challenges of machinery intelligent fault diagnosis in big data era, and then present recent research progress of the speaker's group: (1) A deep learning-based method is presented which could adaptively mine fault characteristics and automatically implement intelligent diagnosis. (2) A two-stage learning method is developed to learn features from raw vibration signals using artificial intelligent techniques as well as classify the mechanical health conditions. These methods will make health monitoring less dependent on human labor than ever when processing mechanical big data.


Prof. Piero Baraldi
Energy Department, Polytechnic of Milan

Subject: "Prognostics and Health Management in the Energy Industry "

Abstract
PHM provides information for effective condition-based and predictive maintenance for increasing productivity, optimizing operating performance, reducing lifecycle costs, extending operating periods between maintenance and reducing downtimes, frequency and severity of unanticipated failures.
As energy is directly related with all industrial activities, improving the reliability of energy delivery while reducing costs is a key issue in the global economy. Inevitably, then, the use of PHM has attracted great interest from industries involved in the production, transportation, distribution and sale of energy.
This keynote lecture will present the specific desiderata of PHM systems for the energy industry. This will be done with reference to the typical problems addressed by PHM, i.e. detecting incipient failures, classifying their causes, predicting the system Remaining Useful Life (RUL), with its corresponding uncertainty, and considering the exploitation of the PHM outcomes for performing condition-based and predictive maintenance. By way of examples of application, the main challenges towards the deployment of PHM systems and their effective integration in the operation and maintenance of energy industries will be discussed.


Prof. Chen Yunxia
School of Reliability and Systems Engineering, Beihang University

Subject: "New challenges of reliability technology in New Energy Battery Industry "

Abstract
The development of new energy battery is the essential way to alleviate crisis of energy scarcity and environment pollution. The key point is to guarantee enough mileages per discharge cyclesafely during the whole lifetime. With restriction of weight and volume, some strict requirements for battery power supply system have been imposed, such as lighter, smaller, higher energy density etc. It brings new challenges of reliability technology including: (1)The utilization of new materials and environment evolving bring new challenges of failure mechanism determination; (2)The Fault diagnosis and prediction process brings new challenges of detection technology; 3)Accurate life prediction brings new challenges of data processing and modeling technology.


Mr. Russell Morris
Boeing Company

Subject: " Impact of PHM on Systems Reliability and Life Cycle Cost "

Abstract
Much of the current work in PHM is to identify the signatures of equipment such that a trend toward failure can be detected early and the part removed from service and replaced before a failure occurs.  Although there is potentially great value in this, there is also a serious drawback as PHM has a significant impact to life cycle cost, spares, and the associated costs of maintenance. This discussion presents some of the challenges and trades that must be considered for PHM to be successful and focuses on attributes that must be considered for PHM to be successful.  These include areas such vibration, shock, wear/fatigue and aging associated with electromigration and chemical degradation. All of these aspects reduce reliability, but the addition of capabilities to measure these effects also affect reliability and can result in false positive as well as false negative detection of failure trends.  In the mechanical domain there are additional issues with the effects of repair on the system failure trend signature. This paper also assesses the impact to the life cycle cost (LCC) of equipment monitored by PHM.

Organizers and sponsors for the conferences: