Real Time Assessment of Novel Predictive Maintenance System based on Artificial Intelligence for Rotating Machines
Bouyahrouzi El Mahdi, El Kihel Ali, El Kihel Youssra, Soufiane Embarki
Abstract
Predictive maintenance (M4.0) allows more targeted and efficient use of resources, reduces unplanned downtime, and increases production and equipment performance compared to classical existing maintenance (M3.0). This paper deals with the development of a new ecosystem that adopts the new technologies of Industry 4.0 to drive real-time monitoring and diagnosis of engine defects. The proposed architecture is based on implementing a process of identifying critical components and extracting related data (speed and acceleration) based on IoT technology. A neural model (ANN) is implemented for monitoring, detecting and diagnosing engine faults with high accuracy compared to existing techniques. The effectiveness and reliability are validated through real-time test bench studies.