Designing Remaining Useful Life (RUL) Algorithms for Predictive Maintenance

Overview

All physical systems degrade over time, and without maintenance they will eventually fail – but when? Most of today’s engineering components collect time series sensor data throughout their lifetime, and engineers can harness this historical data to predict when equipment will fail. This webinar is for engineers who work with time series data and want to accurately predict the Remaining Useful Life (RUL) of equipment to enable predictive maintenance. In this webinar we will explore various types of historical sensor data, interactive feature engineering and condition indicator design, and different types of RUL models using real-world examples in MATLAB and Simulink. 

Highlights

• Exploring, organizing, and preprocessing time series sensor data
• Iterative feature engineering and health indicator design using interactive apps
• RUL estimation models including similarity, degradation, and survival models
• AI approaches to RUL estimation
• Deploying RUL models to edge or cloud environments