The design of a longitudinal smart predictive maintenance system for automotive components using MATLAB is presented. Advantages of predictive maintenance are emphasized, while Remaining Useful Life (RUL) estimation, Prognostics and Health Management (PHM), and Digital Twins are considered as key concepts. Methods are sought to integrate Machine Learning (ML) algorithms into forecasting, a methodology developed for predictive maintenance model development of a quarter car suspension using simulated fault data, and of an automotive engine using public datasets. The paper presents and discusses results, and later concludes with the areas of the benefits of the method in terms of predicting and improving the target performance of the models.