AI-Driven Predictive Maintenance: Deep Learning for Mechanical Parts Life Estimation and Health Monitoring

Authors- Mrs. M. Mani Deepika., M.Yaswanth., K. Abhishek, CH.Rohan., J.Sanjana., A. Pavani

Abstract-To enhance the accuracy of predictions and enable real-time monitoring of mechanical parts, a deep learning-based approach was developed. First, a Convolutional neural network (CNN) was designed to extract key features from mechanical components. These extracted features were then processed through a fully connected layer for information fusion and classification, enabling precise life prediction and health status monitoring. The trained deep learning model was subsequently integrated into a monitoring system, forming a comprehensive framework for mechanical parts’ life prediction and condition assessment. Continuous optimization and updates were performed to improve prediction accuracy, real-time responsiveness, and adaptability to varying working conditions and environmental factors. Experimental results demonstrated that the model achieved a Mean Absolute Error (MAE) of 2.1, a Root Mean Squared Error (RMSE) of 2.5, and a Mean Absolute Percentage Error (MAPE) of 10%. These findings highlight the model’s excellent performance and its potential to provide significant technical support for engineering applications.

DOI: /10.61463/ijset.vol.13.issue2.265