The paper presents the development of engine health monitoring systems for turbofan engines within the aviation domain and their contribution toward predictive maintenance strategies. The methodology elaborated in the paper makes use of permutation entropy for the selection of critical sensors, generates high-quality synthetic sensor data using generative adversarial networks (GANs), and performs anomaly detection using autoencoders. The framework thus trains the machine learning models to enhance the prediction capability under conditions scarce with data. Important contributions include the combination of GANs to render the synthetic data both realistic and reliable in order to make the system more robust and the deployment of autoencoders to fortify anomaly detection and predictive maintenance capabilities. This will entail reduced system downtime and maintenance costs while at the same time increasing the efficiency of predictive maintenance operations. The paper also fills in the whole architecture of the system, implementation, and initial results to demonstrate how it has potential into uprooting conventional maintenance strategy used in the aviation sector and improving operational reliability.