This work focuses on properly estimating lithiumion battery Remaining Useful Life in renewable energy storage and electric vehicle applications. This is important since batteries are essential to sustainable energy systems and electric car demand is expanding. Accurately predicting Remaining Useful Life (RUL) is essential for maximizing battery safety and efficiency. Furthermore, these predictions play a critical role in ensuring that energy storage systems remain both economically viable and environmentally sustainable. The study uses a dataset from the Hawaii Natural Energy Institute, containing data on battery lifecycle performance metrics. The machine learning models used include random forests, decision trees, linear regression, K-nearest neighbors (KNN), support vector machines (SVM), AdaBoost, and extreme gradient boosting. To maximize predictive precision, every model undergoes a process of hyper-parameter tuning, data preprocessing, and the evaluation of feature importance. Comparative measures like RMSE, MAE, and R² assess model predictive capability. The research shows that Random Forest and XGBoost models excel, with Random Forest earning an RMSE of 23.26 and an R2 score of 99.68. XGBoost demonstrated strong performance, including a decreased RMSE of 20.64 and an R² score of 99.61. Research indicates that ensemble learning techniques, such as XGBoost and Random Forest, provide highly reliable forecasts for Remaining Useful Life, facilitating improved maintenance and oversight of battery systems. The novelty of this study lies in its comprehensive comparative analysis of multiple machine-learning models optimized with metaheuristic techniques for battery Remaining Useful Life predictions. In contrast to earlier research that typically utilized single-model methods, this study investigates an ensemble of machine learning models with a focus on optimization and feature importance to enhance predictive precision. The results show that feature selection, data processing, and hyperparameter optimization have improved the model’s performance. This supports machine learning’s role in preventive maintenance, battery longevity, and sustainable energy. This study introduces a robust, data-centric framework that combines various techniques to enhance remaining useful life predictions, serving as a beneficial resource for practical battery management systems.