Background <p>Predictive maintenance strategies for industrial machinery must evolve to address the limitations of traditional vibration-only monitoring systems, particularly their inability to capture spatial context or provide visual validation of mechanical conditions.</p> Methods <p>This study presents an innovative hybrid approach integrating vibration analysis with advanced computer vision techniques, including Zhang’s camera calibration method. The research focuses on real-time monitoring and analysis of machinery vibrations in both time and frequency domains, complemented by computer vision to detect deviations in equipment performance. An experimental setup involving conventional lathes validated the method.</p> Results <p>The hybrid approach demonstrated high effectiveness in identifying potential failures, enabling timely preventive and corrective maintenance. Results showed significant improvements in operational reliability and machinery lifespan, while substantially reducing costs associated with unexpected failures and production downtime.</p> Conclusion <p>This work contributes to the field of vibration engineering by offering a robust, multi-modal framework for condition monitoring and fault detection in industrial settings, providing a scalable solution that enhances predictive maintenance capabilities.</p>