Reliability and uptime are critical priorities in the automotive industry, prompting a shift toward predictive maintenance (PdM) to minimize unexpected failures and associated costs. This study presents a machine learning-based framework for early prediction of engine fuel system failures using embedded field performance data. This study introduces a machine learning-based framework for predicting failures and estimating the remaining useful life (RUL) of mid-range diesel engines with high-pressure common rail fuel systems in vehicles using classification and regression models applied to embedded field performance analysis data, aiming to enhance reliability and reduce unplanned downtime. Two classification models --- Random Forest and XGBoost top our model metrics chart. They were further tuned and evaluated, with XGBoost achieving superior performance, including 94% accuracy and 87% precision, and a low false positive rate of 0.01, enabling an 8-day lead time for proactive maintenance. Additionally, a Gradient boosting regression model was used to predict RUL in miles, achieving R2 values of 0.997 and 0.999 for test and validation datasets respectively, and RMSE as low as 77 and 52 miles for test and validation datasets respectively. Acknowledging challenges such as data quality, computational complexity, and architectural limitations, this study underscores the advantages of our innovative approach and identifies promising avenues for future research in predictive maintenance within the automotive sector, ultimately contributing to increased efficiency and reduced operational costs. Future work will focus on expanding the dataset, enhancing model generalization, and exploring edge computing for real-time deployment.