Industry 4.0 Technologies for an Observer-Based Gearbox Fault Detection Architecture

This paper presents a MQTT-based monitoring architecture for fault detection and diagnosis in industrial electromechanical systems. The proposed system integrates several modules that work in concert to achieve real-time data acquisition, processing, and diagnostic evaluation. In our implementation, a simulation platform is developed in LabVIEW to verify the functionality of the overall architecture. The system comprises a simulated plant that replicates the behavior of an induction motor coupled with a single-stage spur gearbox, an edge computing module that preprocesses measurements and estimates electrical torque, and a diagnostic filter that generates residual signals for fault detection. Communication among these modules is enabled by an MQTT broker hosted on a Raspberry Pi, facilitating efficient real-time data exchange over a wireless network. Experimental evaluations confirm that the proposed architecture is effective in detecting fault-induced changes, underscoring its potential for practical industrial applications.