The widespread adoption of electric vehicles (EVs) has significantly increased the demand for reliable and efficient Battery Management Systems (BMS), which are essential for ensuring safety, high performance, and extended battery life. BMS rely on parameters such as voltage, current, temperature, and charge-discharge cycles to estimate the battery’s State of Charge (SOC) and State of Health (SOH). However, efficiency degradation and fault conditions remain key challenges. This study presents a machine learning (ML)-based framework aimed at detecting efficiency-related anomalies in BMS data. The proposed approach emphasizes a structured preprocessing pipeline, incorporating missing value imputation, outlier correction, skewness handling, and feature scaling to enhance data quality. Unsupervised anomaly detection using Isolation Forest is employed to identify deviations from normal behavior, followed by classification of these anomalies using supervised ML algorithms: Logistic Regression, Decision Tree, and Random Forest. Model performance is rigorously evaluated using Accuracy, F1-Score, and AUC-ROC metrics to benchmark effectiveness. This work contributes a scalable and data-driven diagnostic methodology for intelligent battery systems, addressing real-time fault detection in evolving EV ecosystems.