Passive Optical Vibration Monitoring for Machine Health in Cyber-Physical Systems

Research Topics: Signal Processing, Cyber-Physical Systems, and Intelligent Monitoring. Goals/Objectives: To develop a passive, sensor-free surveillance system that uses CCTV video feeds to analyze minute vibrations in objects to keep an eye on the health of machines. Methodology: To extract vibration signals from live or recorded videos, the suggested system, called Silent Guardian, employs pixel-level intensity tracking. Signal graphing and frame analysis are done with MATLAB. Without the need of physical sensors, the technique enables the early warning detection of mechanical anomalies based solely on visual input. The system is presently being developed and is moving closer to being deployed in real time. Results and Analysis: Based on preliminary testing, Silent Guardian is able to distinguish between typical and abnormal machine conditions by monitoring changes at the pixel level. By using current camera sets to identify early indications of anomalous vibration, the system has the ability to avoid failures.