One of the unique features of the PHM conferences is free technical tutorials on various topics in health management taught by industry experts. 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.
Tuesday, 1:30 PM – Anomaly Detection
Neil Eklund, Analatom
Abstract: Anomaly detection (AD) is the process of identifying elements in a data set which differ from the norm. AD is difficult, because it is typically performed in an unlabeled context, taking only the internal structure of the dataset into account. Anomalous data might be the most interesting data in a particular data set, or it might represent garbage; and can arise from a plurality of sources – noise, stuck sensors, a different underlying system, an unusual mode of operation, and so on. This Tutorial will focus on the concepts and algorithms of AD for PHM data, with real-world examples and advice on algorithm selection for practical tasks.
Bio: Dr. Eklund is an experienced technologist in the space of data science, industrial analytics, and machine learning, with over 20 years of experience in developing fielded solutions to practical industrial problems. He is one of the founders of the PHM Society, and continues to serve on its board of directors. Neil is the former Chief Data Scientist for Schlumberger, where he established the first successful deployed IoT application in the oil industry, which generated $20MM+ in the first three months of operation. Prior to that, Neil was a research scientist in the Machine Learning laboratory of GE Research, working in aerospace, energy, healthcare, oil & gas, financial, and rail applications.
Tuesday, 3:30 PM – Systems Engineering for PHM
Ravi Rajamani, drR² consulting
Abstract: With the goal of delivering predictive maintenance and continuous remaining useful life estimates, PHM systems can reduce aftermarket costs by 25% or more and increase operational availability by 15% or more. This has been proven over many years of experience in the aerospace sector. However, developing PHM systems is not easy because they cut across many different disciplines and subsystems. An SE approach has been shown to be a very good way of designing and implementing PHM systems. This tutorial will give an overview of how SE can be used to develop PHM systems in a systematic manner.
Bio: Dr. Ravi Rajamani is an independent consultant who has accumulated years of experience in the area of aerospace propulsion and energy, specifically in data analytics and model-based methods for controls, diagnostics, and prognostics. He has many publications to his name including three books (chief being Electric Flight Technology: The Unfolding of a New Future), book chapters, journal and conference papers, and patents. In the past Ravi has worked at Meggitt, UTC, and the GE. He is active within various SAE technical committees dealing with PHM. He is also active in the PHM Society, serving on its board of directors. He is the editor-in-chief of the SAE International Aerospace Journal and has been elected a Fellow of SAE and IMechE.
Thursday, 9:00 AM – Prognostics
Marcos Orchard, University of Chile
Abstract: Uncertainty and risk management holds the key for a successful penetration of health management strategies in industrial applications. While methods to estimate and handle uncertainty have received a reasonable amount of attention in the diagnostics domain, uncertainty management for prognostics is an area which awaits major advances. This tutorial explores some of the most important theoretical and practical aspects associated to the problem of failure prognosis and risk evaluation, with special emphasis on Bayesian prognostic algorithms, outer feedback correction loops, and uncertainty characterization in long-term prediction.
Bio: Marcos Orchard is Associate Professor with the Department of Electrical Engineering at Universidad de Chile and was part of the Intelligent Control Systems Laboratory at The Georgia Institute of Technology. His current research interest is the design, implementation and testing of real-time frameworks for fault diagnosis and failure prognostics, with applications to battery management systems, mining industry, and finance. His fields of expertise include statistical process monitoring, parametric/non-parametric modeling, and system identification. His research work at the Georgia Institute of Technology was the foundation of novel real-time fault diagnosis and failure prognostic approaches based on particle filtering algorithms. He received his Ph.D. and M.S. degrees from The Georgia Institute of Technology, Atlanta, GA, in 2005 and 2007, respectively. He received his B.S. degree (1999) and a Civil Industrial Engineering degree with Electrical Major (2001) from Catholic University of Chile. Dr. Orchard has published more than 50 papers in his areas of expertise.
Thursday, 10:45 AM – Deep Learning for PHM
Gabriel Michau, Zurich University of Applied Sciences (ZHAW)
Abstract: Developments in Neural Networks have brought over the past decade a variety of solutions to problems in many fields. Neural Networks are now able to classify, recognize or segment thousands of objects in images, play games, generate new engineered products, translate languages, recognize and translate voices or gestures. Yet, in PHM, the use of neural networks is still quite limited. In this tutorial, we will first explore the fundamentals of deep learning and of its most successful applications. Understanding the strength and the limitations of Neural Networks will help us to understand why their application to PHM is not straightforward. Toy examples will be used to help familiarize with these concepts.
Bio: Dr. Gabriel Michau is a Research Associate in Predictive Maintenance at the Zurich University of Applied Sciences (ZHAW). He holds the joint PhD degree from the Queensland University of Technology (QUT), Brisbane, in Civil Engineering, and from the Ecole Normale Supéerieure de Lyon (ENSL), in Physics, Signal Processing. His research focus is now on machine learning, signal processing and data-driven approaches, applied to the Prognostic and Health Management field.