Predictive maintenance lets you estimate the optimum time to do maintenance by predicting time to failure of a machine. This way, you can minimize downtime and maximize equipment lifetime. In this series, you’ll learn how predictive maintenance works and how it is different from other strategies such as reactive and preventive maintenance. The videos will also walk you through a workflow that will help you develop a predictive maintenance algorithm. You’ll learn about condition indicators and how you can extract them from your data to discriminate between healthy and faulty states. Machine learning models are trained using the extracted condition indicators to classify different types of faults. The videos will also help you understand different estimator models, such as survival, similarity, and degradation, that are used to estimate the remaining useful life of a machine.
Part 1: Introduction Learn about different maintenance strategies and predictive maintenance workflow. Predictive maintenance lets you find the optimum time to schedule maintenance by estimating time to failure.
Part 2: Feature Extraction for Identifying Condition Indicators Watch this video to learn how you can extract condition indicators from your data. Condition indicators help you distinguish between healthy and faulty states of a machine.
Part 3: Remaining Useful Life Estimation Predictive maintenance lets you estimate the remaining useful life (RUL) of your machine. Explore three common models to estimate RUL: similarity, survival, and degradation.
Part 4: How to Use Diagnostic Feature Designer for Feature Extraction Learn how you can extract time-domain and spectral features using Diagnostic Feature Designer for developing your predictive maintenance algorithm.