Improving Electric Vehicle Maintenance by Advanced Prediction of Failure Modes Using Machine Learning Classifications

This study presents a novel approach for analyzing failure modes and their effects in electric vehicles, offering an alternative to traditional methods. We employed machine learning techniques, enabling a significant shift in predictive maintenance. Various models, including Random Forest, Decision Trees, Logistic Regression, and Neural Networks, were tested, with Random Forest and Neural Networks achieving an impressive accuracy of 96.67%. This advancement enhances fault prediction, reduces operational costs, and minimizes downtime by integrating numerical and categorical data. The innovation of this study lies in combining multiple models to detect more complex patterns. Unlike previous studies that relied on simpler methods like decision trees or regression analysis, our multi-layered approach improves prediction accuracy by utilizing real-world data from electric vehicle sensors. This approach overcomes the limitations of simulated data and isolated models, significantly enhancing vehicle reliability, lifespan, and maintenance.