A Cyber-Physical System Approach for Predictive Maintenance

Abstract

Cyber-Physical System (CPS) is considered as one of the main enablers for Industry 4.0. CPS technology integrates knowledge from multiple domains, and bridges the gap between the Physical and the Cyber worlds. Predictive Maintenance (PdM), a crucial application within Industry 4.0, can take a CPS-based approach with intelligent operations to minimize machine downtimes and associated costs. This paper discusses challenges of PdM implementation and evaluates performance of the Convolutional Neural Network (CNN) algorithm in fault diagnosis of a PdM system.