This article proposes a Condition-Based Maintenance (CBM) approach for aircraft engines and Remaining Useful Life (RUL) monitoring, and failure prevention. Due to the unavailability of run-to-failure data, Turbofan Engine Simulation data, obtained from NASA repository, is used to train and test our model. Data Acquisition and Management system framework and planning are proposed for online monitoring and RUL prediction. In practice, sensor measurements usually suffer from noise contamination, hence the prediction models are challenged by noise contaminated data for both training and testing tasks. This is done to assess their prediction ability in a similar condition of having noisy data. Linear and nonlinear prediction models are developed, with performance comparison addressing both regression and classification problems. Models performance indices consider both prediction accuracy and percentage of predictions before the actual failure (PBAF). The proposed model considers continuous learning and improvement to account for any further operational changes that affect the model prediction ability. This is reached by ingesting the model with the actual RUL during the maintenance of the engine unit, and by comparing it to the predicted one.