A Rule-Based Vibration Analysis and Diagnostic (VAD) Platform for Rotating Equipment Maintenance

Abstract

Vibration analysis remains one of the most reliable techniques for monitoring rotating equipment health, yet in most plants it still depends on manual spectrum interpretation by a small pool of specialists, creating bottlenecks and inconsistent diagnoses. This paper presents the Vibration Analysis and Diagnostic (VAD) platform, a rule-based decisionsupport tool that standardizes and accelerates first-level fault diagnosis for pumps, compressors, and turbines. Built in Python with a Tkinter interface, NumPy/SciPy for frequency-domain calculation, Matplotlib for visualization, and FPDF for reporting, the platform applies diagnostic rules drawn from the Mobius Institute vibration guide. Technicians enter equipment specifications and the dominant observed frequency and receive a structured report of probable fault causes and recommended actions, while the platform simultaneously builds a verified historical database of confirmed cases. This paper describes the platform's architecture, operational workflow, and expected benefits in diagnostic speed and consistency, and outlines a roadmap toward a machine-learning-based diagnostic model.