Special Sessions

#1 PHM for Railway Systems through Collection and Processing of Multi-Source Signals

The railway system is an important transportation network in our daily life. Due to the complexity of its mechanical, electrical and hydraulic structures, the state monitoring information of railway system is often a complex signal with multiple sources and heterogeneity. How to collect and process multi-source signals at high speed will directly affect the real-time monitoring and health management of railway system.

This session covers the PHM research of Railway systems with multi-source signals, and welcomes papers with PHM related topics in, but not limited to: synchronous acquisition of multi-source signals, information fusion, feature extraction, feature selection, intelligent maintenance, intelligent dispatching, etc.

Haidong Shao, Hunan University, China. Email
Haiyang Pan, Anhui University of Technology, China. Email
Zhiyi He, Guangzhou University, China. Email

#2 Optimization Problems Related to System Reliability and Maintenance

This session focuses on techniques to evaluate and optimize the reliability of system, considering different system structure and requirements, diagnose the system quality, and make optimal decisions on system maintenance strategy. The presenters are from different universities, different cities, and are of different backgrounds. In addition, this session contains both theoretical models and real case studies. The methodology involves quality assessment chain, Markov chain, combinatorial model, intelligent optimization techniques, etc. We are sure that the audience will find this session interesting and fruitful.

Dr. Kaiye Gao, Beijing Information Science and Technology University, China. Email
Dr. Rui Peng, Beijing Information Science & Technology University, China. Email

#3 Measurement, Diagnosis, Modeling, and Prognostics Methods in PHM

The last decade has witnessed a growing research interest on various aspects of Prognostics Health Management (PHM). PHM plays an important role in maintenance scheduling, maintenance strategy selection, inspection optimization, and spare parts provision. Thus, it can be applied in a variety of areas such as transportations, energy systems, and aerospace. The main focus of this special session is on new theories and methodologies and their applications in measurement, diagnosis, modeling, and prognostics.

Potential topics include but are not limited to:

Yongbo li, Northwestern Polytechnical University, China. Email
Ni Li, Northwestern Polytechnical University, China. Email
Zhiqiang Cai, Northwestern Polytechnical University, China. Email
Zhixiong Li, University of Wollongong, Australia. Email

#4 Dynamic Time Series Modeling, Prediction and Applications in Industrial PHM

In industrial scenarios, the collected health condition signals are very common and popular in the form of time series, such as vibration signals, acoustic signals, so on. Dynamic modeling of such kinds of signals, for many practical applications, is the essential step which enables to analyze their dynamical behavior and characteristics in the past, and predict their trend in the future. Dynamic analysis and modeling of time series, together with short-term/long-term prediction and anomaly detection, covers a wide range of PHM applications in industries, such as transient stability assessment, early fault warning, remaining useful life evaluation, and many other online health management problems. This special session welcomes papers and presentations on various industrial PHM applications with a special focus on dynamic modeling and prediction of time series.

Dr. Guoliang Lu, Shandong University, China. Email
Dr. Wanqing Song, Shanghai University of Engineering Science, China. Email
Dr. Wentao Mao, Henan Normal University, China. Email

#5 PHM based on Digital Twin

A digital twin is a virtual model of a process, product or service. This pairing of the virtual and physical worlds allows analysis of data and monitoring of systems to head off problems before they even occur, prevent downtime, develop new opportunities and even plan for the future by using simulations. Smart components that use sensors to gather data about real-time status, working condition, or position are integrated with a physical item. This data help researchers build more accurate model and make better decisions. Digital Twin allows us to operate, maintain, or repair systems when we aren’t within physical proximity to them.

This session includes but not limited to: complex equipment modeling, complex equipment simulation, fault injection simulation, digital twin algorithm, PHM based on digital twin for aircraft, PHM based on digital twin for train, PHM based on digital twin for any other complex equipment etc.

Guigang Zhang, Institute of Automation, Chinese Academy of Sciences, China. Email
Wenjuan Du, Institute of Automation, Chinese Academy of Sciences, China. Email
Tengfei Zhang, Institute of Automation, Chinese Academy of Sciences, China. Email

#6 Industrial AI Driven Digital Twins for PHM

Intelligent prognostic and health management (PHM) are essential for just-in-time maintenance, which guarantees high-quality products, minimizes unplanned downtime, and increases customer satisfaction. As one of the enables, Industrial Artificial Intelligence (AI) is a systematic discipline to systematically develop and deploy AI algorithms with repeating and consistent successes. Nowadays, driven by Industrial AI tools, digital twins are under developing for not only digitalized operation condition of assets, but also for diagnostic, prognostic and prescriptive analytics in PHM scenarios. Moreover, it helps to move forward to our goal of cognitive analytics (intelligent autonomous actions) in maintenance.

This special session solicits papers that present various PHM program through digital twins driven by Industrial AI, with special focuses on: 1) digital twin architecture and technology that enables convergence of OT, IT and ET; 2) data-driven, model-based, hybrid-driven, as well as context-driven approaches; 3) studies cover from component to System of System level. Among them, techniques and approaches used, results obtained, and lessons learned can be included to share experience with this session.

Baoping Cai, China University of Petroleum (East China), China. Email
Yiliu Liu, Norwegian University of Science and Technology, Norway. Email
Janet Lin, Luleå University of Technology, Sweden. Email

#7 PHM for Thermodynamic System Operation and Maintenance

The dynamic characteristics, diagnosis and control of thermodynamic systems are important research directions of engineering thermodynamics, and are important basic disciplines to realize the safe and economic operation of thermal systems. With the rapid development of various large-scale and complex energy power systems, the extensive application of thermodynamic systems under extreme conditions (high speed, high temperature, high heat flow, and high intensity combustion), and the huge pressure on energy saving and emission reduction, it is becoming more and more urgent to vigorously carry out research on application foundation and key technical issues in the aspects of new modeling theory of thermodynamic system, energy-saving operation optimization and control methods, and system-level fault detection and diagnosis methods. There is an urgent need to integrate scientific research forces in this field to promote the development of disciplines and make due contributions to national economic development.

Dr. Yulong Ying, Shanghai University of Electric Power, Shanghai, China. Email
Dr. Jingchao Li, Shanghai Dianji University, Shanghai, China. Email

#8 Reliability of Intelligent Hybrid Information System

Mobile Information/signal processing has been acted as an important research domain for a long time. In long-distance transportation, themobile traffic control system is high depended on effective and timely information transmission; and the intelligent monitoring system is also required more effective anomaly detection. In general mobile world, which includes industrial, agricultural, and other environment, there are also many information-based problems need to be focus on. Therefore, it is the right time to research on the high reliability of worldwide intelligent hybrid information system. For example, the reliability of technical solutions of information preprocess and systemization, objective tracking, equipment patrol and behavior understanding with sensors in IoT, restoration and enhancement of interesting content of information, et al.

Prof. Shuai Liu, Hunan Normal University, China. Email

#9 PHM for Wind Turbine Operation and Maintenance

Wind turbine is a comprehensive product integrating electrical,mechanical, air mechanics and other disciplines, all parts of which are closely related. Successful PHM will be a must for more effective maintenance, increasing capacity, and more reliable and robust for existing wind energy system. The quality and maintenance efficiency of each component of the Wind turbine product also directly affect the operation and maintenance level of the Wind turbine. This session covers the PHM research of Wind turbine operation and maintenance, and welcomes papers with PHM related topics in, but not limited to: Life prediction of parts, wind speed forecasting, dynamics, signal processing, condition monitoring, state detection, fault diagnosis and prognosis, health management, maintenance strategy, etc.

Baoping Tang, Chongqing University, China. Email
Zhipeng Feng, University of Science and Technology Beijing, China. Email
Zong Meng, Yanshan University, China. Email
Ling Xiang, North China Electrical Power University, China. Email

#10 Advances in Prognostics and Health Management for Intelligent Manufacturing

As an emerging field in the Mechanical Science, Industrial Engineering (IE) and Prognostics and Health Management (PHM) is gaining interests from the industry and academia. An effective PHM framework normally includes health prognostics and maintenance management. Meanwhile, advances in sensor technology and wireless communication play a vital role in enabling Industrial Internet of Things (IIoT). Modern technologies, such as Industry 4.0 and cyber-physical systems, can improve the capability of health prognostics and the accuracy of maintenance management.

Mechanical failures are the “potential killer” for the safe and reliable operation of large-scale mechanical equipment, such as offshore wind farms, aviation engines and CNC machines. And health prognostics is the “trump card” to ensure the normal operation of mechanical equipment. Recently, more and more professional and hi-tech instruments (e.g. smart sensors, meters, controllers and computational devices) have been applied to collect and analyze the signals from individual machines. Prognostics techniques, such as vibration monitoring, oil analysis, temperature detection, acoustic emission and ultrasonic inspection, have been widely employed to measure the status of machines. Many valuable prognostics approaches have thus been proposed to generate rational estimation and prediction of remaining useful life (RUL) or potential degradation process.

Moreover, maintenance policies of complex systems are facing challenges from structural, stochastic, economic dependencies and advanced manufacturing paradigms. Due to recent developments in the manufacturing paradigms, PHM methodologies for traditional manufacturing systems need to be extended. Specially, contents in PHM have been studied a lot by integrating prognostics techniques with maintenance policies, such as condition-based maintenance (CBM), predictive maintenance (PdM) and on-condition maintenance (OM). Therefore, machine interactions and production characteristics should be investigated and modeled to identify maintenance opportunities for achieving a cost-effective and timely maintenance scheme.

The aim of this Special Session is to promote prognostics and health management, and act as a platform to present high-quality original research on the latest developments of health prognostics and maintenance management for intelligent manufacturing. We welcome both original research articles and review articles discussing the current state of the art.

Potential topics include but are not limited to the following:

Dr. Tangbin Xia, Shanghai Jiao Tong University, China. Email
Dr. Dong Wang, Shanghai Jiao Tong University, China. Email

#11 Reliability and Resilience for Networked Systems

A challenging issue in reliability and resilience area is that the scale and complexity of various practical systems increase quickly along with the rapid popularization and development of the Internet technology, which make it more difficult to evaluate and predict the system state. For these systems such as the Internet, the critical infrastructure system, the manufacturing system and so on, the classical reliability and resilience modeling methods might no longer be applicable and new modeling approaches are needed. This session focuses on this issue

Dr. Ning Wang, Chang’an University, China. Email
Dr. Dongli Duan, Xi’an University of Architecture and Technology, China.

#12 Industrial Big Data Mining and Artificial Intelligence for PHM

With the advance of Industrial Internet platform and the upcoming application of 5G communication infrastructure, the Industrial big data mining and artificial intelligence technologies for prognostics and health monitoring (PHM) of machinery, is the focus of the next stage. However, the Industrial big data in Internet of Things is with a certain redundancy,complex and unstructured characteristic, it is a great challenge for analyzing and processing industrial big data in the application of PHM. Recently, artificial intelligence such as deep learning and transfer learning has achieved great success on the value mining of the industrial big data, which include the discovery of new patterns and knowledge and the extraction of novel valuable information. This special session solicits papers that present various industrial PHM applications of machinery such as bearing, gearbox and hydraulic device, with a special focus on industrial big data mining and artificial intelligence technologies. These include but not limited to: intelligent algorithms and modelling for anomaly detection, fault diagnosis and remaining useful life prediction of machinery.

Prof. Jun Wu, Huazhong University of Science and Technology, Wuhan, China. Email
Prof. Yi Qin, Chongqing University, Chongqing, China. Email
Prof. Yang Liu, Northeastern University, Shenyang, China. Email
Dr. Yuanhang Wang, China Electronic Product Reliability and Environmental Testing Research Institute, Guangzhou, China. Email

#13 Diagnostic and Prognostic Methods under Multi-Source Uncertainty

Multi-source uncertainty arising during the design, manufacturing and operation process significantly challenges the prognostic and health management (PHM) in related communities. Experience in a number of engineering disciplines reveal the necessity to quantify and manage such uncertainties in order to keep project risks in check and system performance accountable. In recent years, enormous studies addressing various uncertainties such as the noise contamination, modelling error and environmental variability are developed to produce more accurate results in a trustworthy level.

Witnessing the promising development, this special session aims at gathering contributions from both academia and industry that discuss new theoretical developments and advanced applications of diagnostic and prognostic techniques involving a multitude of sources of uncertainty.

Potential topics include but are not limited to the following:

Dr. Lechang Yang, University of Science and Technology Beijing, China. Email
Dr. Pidong Wang, University of Science and Technology Beijing, China. Email
Dr. Sifeng Bi, Beijing Institute of Technology, China. Email

#14 Fault diagnosis and Intelligent Prognosis for Mechatronics System

The mechatronics systems are widely used in the complex engineering equipment. The research on the PHM of the mechanics systems the key component of the overall PHM system of the equipment. Fault diagnosis and intelligent prognosis is one of the critical enable techniques of PHM. This session focuses on the recent original research progress in the innovative methods, intelligent techniques and application cases of fault diagnosis and intelligent prognosis for the specific mechatronics systems.

Cheng Zhe, National University of Defense Technology, China. Email
Guanghan Bai, National University of Defense Technology, China. Email

#15 Degradation Modelling and Reliability Analysis for Complex Systems

Degradation modeling and reliability estimation for systems such as MEMS, wind turbines, etc., that experiencing natural degradation and shocks is essential for maintenance decision making and contingency mitigation. A significant amount of research results has been reported in this field. They can be classified into experience-based models, data-driven models and physics-based models. Some of these models consider only one failure mode, e.g., soft failure which is caused by physical deterioration such as wear, erosion, and aging. Other models consider both soft failures and hard failures triggered by sudden fatal shocks in the form of blows, jolts, etc. Nowadays more and more systems or equipment installed in the open air suffer from various environmental shocks, which makes it more complicated for degradation modeling. This invited session aims to bring experienced researchers in this field together to share ideas, results, and applications of degradation and shock modeling.

Jia Wang, Hebei University of Technology, China. Email
Guanghan Bai, National University of Defense Technology, China. Email

#16 Prognostics and Health Management On Complex Mechatronic Systems

Mechatronic systems integrate diversified entities of electrical, electronic, and mechanical components, suggesting complicated failure mechanisms. Safety operation of complex mechatronic systems poses critical impacts on industrial production, resources, and environment. When they are subjected to severer environments like high power, heavy load, the incipient anomaly behavior will result in catastrophic failures if preventive measure is not implemented in advance. The coupling and latent relations between the electrical, electronic, and mechanical sub-systems significantly increase the risk of failure and pose unprecedented challenges to prognostics and health management (PHM) applications. To eliminate the underlying risks posed by those complicated mechanisms, there are urgent demands for information techniques and artificial intelligence (AI) tools for detecting hazardous faults to ensure system safety. This special session covers the PHM applications of mechatronic systems with complex coupling mechanisms. The scope includes but is not limited to:

  1. Accelerating testing experiments of mechatronic components and systems.
  2. Advanced data analytics for condition monitoring of mechatronic systems.
  3. Case studies of detection, fault diagnosis, and prognostics on mechatronic components, e.g., mechanical-electric actuators, drivetrain, avionics, and power electronic devices, etc.
  4. Hardware implementations of PHM prototypes, e.g., edge intelligence applications.
  5. State-of-the-art AI tools for mechatronic systems, e.g., deep learning tools, digital twins, etc.

Shaowei Chen, Northwestern Polytechnical University, China. Email

#17 PHM for Gears

Gear is a common component in mechanical equipment, and gear transmission is also one of the most common ways in mechanical transmission. In many cases, gear failure is an important reason for equipment failure. Therefore gear health management and maintenance is very important.

For the gear PHM, Professor Li Zhinong, who is from Nanchang Hangkong University, will organize a Special Session, which is named as Gear PHM, in the upcoming PHM-Nanjing 2021 conference. Topics of Interest are listed as follows:

  1. Model-based prediction method of gears
  2. Prediction method of gears based on data-driven
  3. Hybrid prediction method of gears
  4. Performance degradation of gears
  5. Reliability assessment of gears
  6. Feature extraction and pattern recognition of gear fault & failure
  7. Other topics related to gear PHM

Welcomes scholars focusing on PHM for gears, to submit your papers to this special session. Your abstract and paper can be submitted to the conference using the following web link: https://easychair.org/conferences/?conf=phmnanjing2021.

Zhinong Li, Nanchang Hangkong University, China. Email

#18 Reliability Modeling and analysis for Supply Chain Networks

Supply chains are the backbone of development of the global economy, thus any disruptions to them are sure to be costly. In reality, supply chain networks that consist of supply, transshipment and demand are susceptible to many unexpected events, such as supplier discontinuities, transportation blockage, natural disasters, power outages, labor strikes, or terrorism, such that the predictability of service delivery may be unavailable. A reliable and robust supply chain network is of critical importance to the stability of economic growth. Actually, the reliability of supply chain networks is the most important aspect of logistics performance. Due to the inherent complexity, reliability modeling and analysis of supply chain networks are regarded to be a rather challenging task. This session welcomes papers with the following topics, but not limited to: Reliability modeling for supply chain networks, algorithms for performance assessment of supply chain networks, reliable and robust supply chain network design, reliability management of supply chain systems, and resilience analysis of supply chain networks.

Yi-Feng Niu, Chongqing University of Posts and Telecommunications, China. Email
Dong Ding, Chongqing University of Posts and Telecommunications, China. Email
Nian Zhang, Chongqing University of Posts and Telecommunications, China. Email

#19 Degradation Process Modeling, Prognostics and Optimization for Manufacturing Systems

The degradation of manufacturing systems can increase the fraction of nonconforming; reduce production rates and even the unexpected breakdown of systems. Therefore, minimizing the effects of degradation in manufacturing systems is of practical importance. This session focuses on the degradation process modeling, remaining useful life prediction, and maintenance optimization of manufacturing systems.

Renyan Jiang, Changsha University of Science and Technology, China. Email
Yifan Zhou, South East University, China. Email

#20 Bayesian Neural Networks for RUL Prognosis

While data-driven techniques and Bayesian approaches have their own advantages in their use for remaining useful life prognosis, data-driven techniques often suffer from the arbitrary choice of a highly complex network architecture which makes the network overfit the data and also the network calibration becomes very sensitive to the initial values of the large number of network weights and biases. On the other hand, Bayesian approaches have their own set of limitations in being incapable of modeling the specific non-ideal features and patterns of the degradation data which are not fully captured by commonly used physics-of-failure models. As such, prognosis can be significantly improved if Bayesian methods (such as particle filters) are combined with data driven neural networks so that the degradation model can be formulated accounting for the complexity of the time series data and at the same time the network weights and biases are learnt through a Bayesian formulation that would be more robust. In effect, the combined use of Bayesian methods and neural networks will enable us to effectively model complex degradation trends and avoid the possibilities of local optima solutions that traditional neural networks often get stuck at. This session welcomes your submissions in the following topics:

  1. Neural Networks inside Particle Filters for RUL Prognosis
  2. Particle Filters inside Neural Networks for RUL Prognosis
  3. Bayesian Approaches to Parameter Training of Neural Networks for PHM
  4. Use of Neural Networks for State-Space Model Formulation in Particle Filters for PHM
  5. Approaches to Optimize the Neural Network Architecture for Degradation Data Prognosis

Prof. Nagarajan Raghavan, Singapore University of Technology and Design (SUTD), Singapore Email

#21 Data-driven Techniques for Rotating Machinery Prognostics and Health Management

Rotating machinery is widely used in industrial applications. Many prognostics and health management (PHM) methods have been developed for rotating machinery using condition monitoring data. However, some challenges still exist. More advanced data-driven techniques are needed to deal with at least the following scenarios: (1) signals contain very weak fault information due to strong environmental noises and non-fault related interferences; (2) signals are strongly non-stationary when machines run at varying speed and load conditions; (3) the remaining useful life is hard to predict due to material and environmental uncertainties existed in the degradation process. This session focuses on data-driven techniques for rotating machinery prognostics and health management. Topics of interest include but not limited to:

  1. Non-stationary signal processing techniques for machinery PHM
  2. Data-driven approaches to amplify weak fault information
  3. Data-driven methods to deal with the material and environmental uncertainties
  4. Condition-based maintenance optimization

Dr. Yuejian Chen, University of Alberta, Canada. Email
Dr. Sajad Saraygord Afshari, University of Manitoba, Canada. Email
Dr. Xihui Liang, University of Manitoba, Canada. Email
Dr. Ming J. Zuo, University of Alberta, Canada. Email