One of the unique features of the PHM conferences is free technical tutorials on various topics in health management taught by industry experts. At PHM16, tutorials will take place on Monday, October 3. As educational events tutorials provide a comprehensive introduction to the state-of-the-art in the tutorial’s topic. Proposed tutorials address the interests of a varied audience: beginners, developers, designers, researchers, practitioners, and decision makers who wish to learn a given aspect of prognostic health management. Tutorials will focus both on theoretical aspects as well as industrial applications of prognostics. These tutorials reach a good balance between the topic coverage and its relevance to the community.
- Model-Based Diagnostics (Slides)
Presented by: Matthew Daigle, NASA Ames Research Center, and Indranil Roychoudhury, SGT Inc, NASA Ames Research Center
- An Introduction to Data-Driven Prognostics of Engineered Systems (Slides)
Presented by: Jamie Baalis Coble, University of Tennessee, Knoxville
- Security Prognostics (Slides)
Presented by: Scott C Evans, General Electric Global Research
- Big Data Analytics (Slides)
Presented by: John Patanian, General Electric Power
|Key Conference Dates|
|Tutorials||3 October, 2016|
PHM Conference tutorials have been a popular event in the past and the PHM society is proud to continue this service to the community. Topics of interest of these tutorials span fundamentals of PHM (Diagnostics, Prognostics, Health Management, Uncertainty Management, etc.) as well as specialized topics such as Cost-Benefit analysis, Data-Mining, Electronics PHM, Bayesian Filtering for Prognosis, etc. For a more comprehensive list of past tutorials please look at the following links:
|Past PHM Tutorials|
George Vachtsevanos firstname.lastname@example.org
Kai Goebel email@example.com
|Tutorial Title: Model-Based Diagnostics (Slides)
Matthew Daigle, NASA Ames Research Center, and Indranil Roychoudhury, SGT Inc., NASA Ames Research Center
|Abstract: The area of diagnostics is focused on the detection, isolation, and identification of system faults. Diagnostics is critical in guaranteeing correct, efficient, and safe operation of complex systems. In model-based diagnostics, faults are diagnosed by reasoning over a model of the system that captures both nominal and faulty behavior. While model-based diagnosis of static systems is well-established, diagnosis of dynamic systems presents a number of additional challenges, and many different approaches have been developed to handle them using different kinds of models and reasoning algorithms. This tutorial will present the general approach of model-based diagnostics, survey different fault diagnosis approaches available in literature, and present a framework for model-based diagnosis of dynamic systems. Advanced concepts of distributed diagnosis will also be presented.|
|Presenter Bio: Matthew Daigle received the B.S. degree in Computer Science and Computer and Systems Engineering from Rensselaer Polytechnic Institute, Troy, NY, in 2004, and the M.S. and Ph.D. degrees in Computer Science from Vanderbilt University, Nashville, TN, in 2006 and 2008, respectively. From September 2004 to May 2008, he was a Graduate Research Assistant with the Institute for Software Integrated Systems and Department of Electrical Engineering and Computer Science, Vanderbilt University, Nashville, TN. During the summers of 2006 and 2007, he was an intern with Mission Critical Technologies, Inc., at NASA Ames Research Center. From June 2008 to December 2011, he was an Associate Scientist with the University of California, Santa Cruz, at NASA Ames Research Center. Since January 2012, he has been with NASA Ames Research Center as a Research Computer Scientist. His current research interests include physics-based modeling, model-based diagnosis and prognosis, simulation, and hybrid systems. Dr. Daigle is a member of the Prognostics and Health Management Society and the IEEE.
Indranil Roychoudhury received the B.E. (Hons.) degree in Electrical and Electronics Engineering from Birla Institute of Technology and Science, Pilani, Rajasthan, India in 2004, and the M.S. and Ph.D. degrees in Computer Science from Vanderbilt University, Nashville, Tennessee, USA, in 2006 and 2009, respectively. Since August 2009, he has been with SGT, Inc., at NASA Ames Research Center as a Computer Scientist. His research interests include hybrid systems modeling, model-based diagnostics and prognostics, distributed diagnostics and prognostics, and Bayesian diagnostics of complex physical systems. Dr. Roychoudhury is a Senior Member of the IEEE and a member of the Prognostics and Health Management Society and the AIAA.
|Tutorial Title: An Introduction to Data-Driven Prognostics of Engineered Systems (Slides)
Jamie Baalis Coble, University of Tennessee, Knoxville
|Abstract:Approaches to prognosis of components and systems are typically divided into model-based and data-driven algorithms. Model-based algorithms rely on first principles based physics of failure models of the evolution of degradation. Data-driven methods use historic run-to-failure and accelerated degradation test data to discover the underlying relationships between measured data and equipment lifetime. Algorithms for data-driven prognostics can be categorized into three types according to the type of information used for prognosis, generally in order of greater specificity and accuracy. Type I (reliability-based) prognostics uses traditional reliability analysis to estimate the lifetime of an average component operating under average conditions. Type II (stressor-based) prognostics incorporate information about how a component or system will be operated (e.g., load, temperature, speed, pressure, demand) to evaluate the lifetime of an average component operating in some specific environment. Type III (degradation-based) prognostics track the condition of a specific component under its specific operation. This condition (or some measure indicative thereof) can be trended to failure.
This tutorial will introduce the general concept of prognostics and place it into context in a full health management system. Empirical prognostic algorithms in each of the three types will be presented.
|Presenter Bio: Dr. Jamie Baalis Coble is an Assistant Professor in the Nuclear Engineering department at the University of Tennessee, Knoxville. Dr. Coble’s expertise is primarily in statistical data analysis, empirical modeling, and advanced pattern recognition for equipment condition assessment, process and system monitoring, anomaly detection and diagnosis, and failure prognosis. Dr. Coble is currently pursuing research in prognostics and health management for active components and systems. Her research interests expand on past work in monitoring and prognostics to incorporate remaining useful life estimates into risk assessment, operations and maintenance planning, and optimal control algorithms. Prior to joining the faculty at UTK, she worked in the Applied Physics group at Pacific Northwest National Laboratory. Her work there focused primarily on data analysis and feature extraction for detecting anomalies and degradation in large passive components (e.g., concrete structures, pipes, welds), advanced active components (e.g., pumps, motors, valves), and other nuclear systems.|
|Tutorial Title: Security Prognostics (Slides)
Scott C Evans, General Electric Global Research
|Abstract: In this Tutorial we cast a vision for Security Prognostics (SP) for critical systems, promoting the view that security related protections would be well served to integrate fully with Monitoring and Diagnostics (M&D) systems that assess the health of complex assets and systems. To detect complex Cyber threats we propose combining system parameters already in use by M&D systems for Prognostics and Health Monitoring (PHM) with security parameters. Combining system parameters used by M&D to detect non-malicious faults with the system parameters used by security schemes to detect complex Cyber threats will improve: (a) accuracy of PHM (b) security of M&D, and (c) availability and safety of critical systems. We also introduce the notion of Remaining Secure Life (RSL), assessed based on the propagation of “security damage,” to create the prospect for Security Prognostics. RSL will assist in the selection of appropriate response(s), based on breach or compromise to security component’s and potential impact on system operation. An example of M&D data is provided which is normally associated with non-malicious faults providing input to detect Malware execution through time series monitoring.|
|Presenter Bio:Dr. Scott C. Evans is Senior Research Engineer in the Machine Learning Lab at General Electric Global Research in Niskayuna, NY. He has 39 patents and over 45 publications in the areas of algorithms, wind analytics, sequence analysis, cyber-security, and wireless network routing / Quality of Service (QoS). Scott holds a PHD in Electrical Engineering from Rensselaer Polytechnic Institute, an MS in Electrical Engineering from the University of Connecticut and a BS in Electrical Engineering from Virginia Tech. Scott is currently a key contributor and machine learning task leader on a $5.6 million IARPA program applying machine learning and causal inference to detect insider threat. Before joining General Electric Research, Scott served as a nuclear-trained Submarine Officer in the United States Navy.|
|Tutorial Title: Big Data Analytics (Slides)
John Patanian, General Electric Power
|Abstract: Big Data is a widely used, perhaps overused term when discussing modern analytics applications. While there is a lot of hype, there are many examples of not previously feasible capabilities enabled by big data technologies, such as large scale exploratory analysis, feature engineering and predictive modeling.
In open source software, Big Data is synonymous with the Apache Hadoop tech stack. The presentation will review key analytics related components of Hadoop including HDFS, Kafka, Hive, Spark, Sqoop, Oozie, and Yarn and their function in batch, interactive, and streaming use cases. Special attention will be given to how analytics have greatly expanded in the transition to Apache Spark and the inclusion of Python and R as first class components.
The tutorial will feature an applied example where Big Data tools were used in developing an anomaly detection algorithm.
|Presenter Bio: John Patanian is Principal Engineer, analytics for GE power and has over 20 years experience in software development, machinery diagnostics, product management, controls optimization, and thermodynamic performance. He holds a masters degree in Computer Science from the University of Washington and a Bachelor’s degree in Mechanical Engineering from Rensselaer Polytechnic Institute. He holds two U.S. Patents and served in the ASME PTC46 committee on performance testing of Combined Cycle power plants.|