A Predictive Maintenance Application for A Robot Cell using LSTM Model. Issue 19 (2022)

• Record Type: Journal Article
• Authors: Joseph, Doyel; Gallege, Tilani; Bekar, Ebru Turanoglu; Dudas, Catarina; Skoogh, Anders
• Is Part Of: IFAC-PapersOnLine. Volume 55:Issue 19(2022)
• Page Start: 115
• Page End: 120
• Publication Date: 2022
• Subjects: Smart Maintenance -- Predictive Maintenance -- Machine Learning -- Long Short-Term Memory (LSTM) -- CRISP-DM -- Industrial Robots -- Manufacturing; Automatic control -- Periodicals; 629.805
• DOI: 10.1016/j.ifacol.2022.09.193
• Languages: English
• ISSNs: 2405-8963

Abstract

Abstract: Maintaining equipment is critical for increasing production capacity and decreasing production time. With the advent of digitalization, industries are able to access massive amounts of data that can be used to ensure their long-term viability and competitive advantage by implementing predictive maintenance. Therefore, this study aims to demonstrate a predictive maintenance application for a robot cell using real-world manufacturing big data coming from a company in the automotive industry. A hyperparameter tuned Long Short-Term Memory (LSTM) model is developed, and the results show that this model is capable of predicting the day of failure with good accuracy. The difficulties inherent in conducting real-world industrial initiatives are analyzed, and recommendations for improvement are presented.