Condition Monitoring of Test Equipment Using Autoencoders
Müller, Mark-Felix (2020)
2020
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2020120225668
https://urn.fi/URN:NBN:fi:amk-2020120225668
Tiivistelmä
This project aimed to research and develop a condition monitoring solution for test equipment of a robotic manufacturing line. Measurement data collected by the test equipment is used to train an artificial neural network to assess the condition of its future measurement performance.
The project involved investigating the measurement data to derive an understanding of the performance of the test equipment. This investigation led to the development of deep learning software used to power a visual display of the results.
Characteristics in the data indicating condition were discovered, which led to the development of an autoencoder used to process measurement data. The processed data is manipulated to determine the condition of the test equipment. The condition of the test equipment is displayed in a graphical view for monitoring.
The result is a monitoring solution which provides timely insight on the tester data that may be used to perform maintenance on the equipment. The project, its research, and its monitoring solution contribute to the development of a predictive maintenance solution.
The project involved investigating the measurement data to derive an understanding of the performance of the test equipment. This investigation led to the development of deep learning software used to power a visual display of the results.
Characteristics in the data indicating condition were discovered, which led to the development of an autoencoder used to process measurement data. The processed data is manipulated to determine the condition of the test equipment. The condition of the test equipment is displayed in a graphical view for monitoring.
The result is a monitoring solution which provides timely insight on the tester data that may be used to perform maintenance on the equipment. The project, its research, and its monitoring solution contribute to the development of a predictive maintenance solution.