Course Overview

This course presents the tools and techniques for development and implementation of prognostics and health monitoring, from both the Physics-of-Failure and the Data Driven perspectives. It  addresses methods for identifying and analyzing precursors based on failure mechanisms, methods for in-situ monitoring, approaches for resource-efficient data collection, algorithms for data reduction and feature extraction, approaches for diagnosing faults, and techniques for failure prediction that can be used for assisting maintenance and logistics decisions. Different approaches for prognostics are presented along with several case studies illustrating the implementation of the methods.

Duration

Seven hours, taught as either a one day or two half-day sessions

Course Outline

1.    Introduction to Prognostics and Health Monitoring

  • Needs and Benefits 
  • Approaches

2.    Monitoring Failure Precursors Based on Physics of Failure

  • Understanding the Life Cycle Profile
  • Failure Modes, Mechanisms and Effects Analysis (FMMEA)
  • Sensors

3.    Life Consumption Monitoring

  • Fuses and Canaries

4.   Data-Driven PHM 

  • Data Pre-processing
  • Anomaly Detection and Clustering
  • Feature Discovery
  • Dimensionality Reduction 

5.    Classification 
6.    Prognosis
7.    Deep Learning for Diagnostics and Prognostics

Related CALCE Courses

  • Physics of Failure of Electronic Products

Contact

Dr. Michael H. Azarian
mazarian@umd.edu
University of Maryland
College Park, MD 20742
 

For more information about CALCE Courses and how to schedule them, Please contact: education@umd.edu 


Top