Anomaly Detection for Industrial Processes and Machinery with MATLAB

Many industries are looking to AI to deliver increased efficiency and improve product quality by automating production process monitoring and maintenance scheduling. Even when production lines are instrumented with sensors as part of digital transformation, engineering teams often lack the specialized skills required by predictive maintenance and advanced process analytics. This webinar will demonstrate statistical and machine learning techniques in MATLAB on real-world datasets to monitor manufacturing processes and detect equipment anomalies. 

Highlights:

- Preprocessing sensor data

- Identifying condition indicators

- Using deep learning and machine learning for anomaly detection algorithms

- Operationalizing algorithms on embedded systems and IT/OT systems

Related Resources:

- Download Example on File Exchange: Industrial Machinery Anomaly Detection: https://bit.ly/46QNxWf

- Anomaly Detection Overview: https://bit.ly/3Re46SO

- Predictive Maintenance Overview: https://bit.ly/3AUp7wR

About the Presenter:

Timothy Kyung is an Application Engineer at MathWorks supporting the Government and Defense Industry with technical expertise in application deployment, interfacing with third party software, and parallelization. He holds a B.S. and M.S. in Mechanical Engineering with a focus in robotics from Carnegie Mellon University. 

00:00 Introduction

02:03 Why do Anomaly Detection?

02:40 What is an Anomaly?

03:53 Anomaly Detection Challenges

06:25 Anomaly Detection Techniques

07:03 Anomaly Detection Algorithm Development Workflow

07:53 Example: Process Monitoring for Copper Production

16:20 Example: Anomaly Detection in Welder Robot Vibration Data

27:27 Deploying Anomaly Detection Algorithms

28:10 Summary