[5D3] Condition monitoring machine learning (ML) model for predictive maintenance of unique manufacturing assets

A Shaalan and D Baglee
University of Sunderland, UK 

Modern manufacturing assets encounter various issues due to being uniquely designed and manufactured for specified products. Such assets do not encounter thorough reliability applications to confirm their applicability to provide high production standards. In addition, maintenance staff knowledge and experience with the asset and the existing maintenance practices that should exist is a vital element to ensure efficient maintenance practices. The current paper examines a unique manufacturing asset’s failure record and examines the applications of different machine learning (ML) models that can support predictive maintenance applications.