IoT-Based Vibration Analysis for Diagnosis of Combined Faults in Industrial Rotary Machinery

Joon, Tzen (2026) IoT-Based Vibration Analysis for Diagnosis of Combined Faults in Industrial Rotary Machinery. Final Year Project (Bachelor), Tunku Abdul Rahman University of Management and Technology.

Text JOON TZEN_Full Text.pdf Restricted to Registered users only Download (16MB)

Abstract

This research develops an IoT-enabled vibration analysis framework for diagnosing single and compound faults in industrial rotary machinery. Using a Machine Fault Simulator (MFS), systematic experiments were conducted on healthy shafts, bearings with ball / inner / outer race defects, bent shafts, and their combinations. High-frequency triaxial sensors captured vibration signals in acceleration, velocity, and displacement, which were analyzed in both time and frequency domains. Results showed clear distinctions between shaft faults (low-frequency sinusoidal distortions), bearing defects (high-frequency impulsive signatures), and compound conditions (chaotic mixed patterns). A comparative database with statistical validation (mean, standard deviation, 95 percent CI) was established, providing a reliable diagnostic reference. This study demonstrates how vibration-based fault signatures, integrated with IoT sensing, can support predictive maintenance and Industry 4.0 practices by enabling early and accurate fault identification in real-world rotating machinery.