Rotary Insight: A Framework for Deep Learning Driven Fault Diagnosis and Health Monitoring of Rotary Machinery, By Amin Torabi Jahromi

Research

Title	Rotary Insight: A Framework for Deep Learning Driven Fault Diagnosis and Health Monitoring of Rotary Machinery
Type	Presentation
Keywords	Bearing fault diagnosis framework, Bearing fault diagnosis, Predictive maintenance, Deep learning
Researchers	Mehdei Tanzadeh (First researcher) , Hossein Haghbin (Second researcher) , Amin Torabi Jahromi (Third researcher) , Seyede zohreh Mousavi (Fourth researcher) , Mohammad Hasan Tavakoli (Fifth researcher) , Hamid Haidarasl (Not in first six researchers)

Abstract

This paper introduces Rotary Insight, an open-source, user-friendly framework for bearing fault diagnosis and health monitoring in rotary machines. The platform provides an intuitive interface for interacting with time-series vibration data, enabling users to upload, segment, and classify faults easily. It automatically preprocesses data, segments it, and provides predictions, including fault classifications and visualizations such as spectrograms. Aimed primarily at educational and research purposes, Rotary Insight also provides a modular environment for experimenting with various models and datasets, making it a valuable tool for predictive maintenance. Future work will focus on improving scalability and knowledge transfer for broader industrial use.