Compact Rotating System Fault Detection*
Session Number
2
Advisor(s)
Dr. Phadmakar Patankar, IMSA
Location
A129
Discipline
Engineering
Start Date
15-4-2026 11:10 AM
End Date
15-4-2026 11:55 AM
Abstract
Predictive maintenance is critical for the efficiency and performance of industrial operations, and while conventional predictive maintenance is often expensive and dependent on cloud infrastructure, the project seeks to design and implement a low-cost fault detection system for rotating machinery using TinyML. The proposed system utilizes a resource-constrained microcontroller (Arduino Nano 33 BLE Sense) to perform real-time monitoring of the health conditions of rotating machinery without the need for external computing infrastructure. In terms of methodology, the proposed system collects and synchronizes both vibration and acoustic signals for the detection of common failure modes such as imbalance, misalignment, bearing defects, and looseness. Through feature extraction using both time and frequency domain analysis (FFT), the proposed system optimizes and implements a neural network for accurate and real-time fault detection. The system is designed to be portable and energy-efficient while accurately detecting faults in rotating machinery, thereby providing a bridge between conventional mechanical vibration analysis and artificial intelligence
Compact Rotating System Fault Detection*
A129
Predictive maintenance is critical for the efficiency and performance of industrial operations, and while conventional predictive maintenance is often expensive and dependent on cloud infrastructure, the project seeks to design and implement a low-cost fault detection system for rotating machinery using TinyML. The proposed system utilizes a resource-constrained microcontroller (Arduino Nano 33 BLE Sense) to perform real-time monitoring of the health conditions of rotating machinery without the need for external computing infrastructure. In terms of methodology, the proposed system collects and synchronizes both vibration and acoustic signals for the detection of common failure modes such as imbalance, misalignment, bearing defects, and looseness. Through feature extraction using both time and frequency domain analysis (FFT), the proposed system optimizes and implements a neural network for accurate and real-time fault detection. The system is designed to be portable and energy-efficient while accurately detecting faults in rotating machinery, thereby providing a bridge between conventional mechanical vibration analysis and artificial intelligence