The object of research is the prediction system for the use of live flight tracker technology on the Boeing 737-900ER aircraft. The problems solved are related to the low accuracy of the prediction system that only relies on technical data without considering aspects of user behavior, as well as the limitations of interpretability in conventional deep learning models that hinder decision validation in critical and sensitive flight environments. The essence of the results obtained is the development of a prediction model based on bidirectional long short-term memory combined with an attention layer and psychological elements from the theory of planned behavior. This model is able to increase prediction accuracy up to 91.2%, much higher than conventional models with an accuracy of around 78%, and shows high F1 and AUC scores indicating a balance between precision and sensitivity. Due to its features and characteristic differences, namely the integration of bidirectional sequential learning, focusing on the most relevant input features through the attention mechanism, and psychological contextualization through the theory planned behavior, these results make it possible to effectively solve the problems of low accuracy and lack of interpretability in predicting flight tracker usage. These results are explained by the model’s ability to highlight key variables such as usage time, flight conditions, and previous interaction patterns that correlate with user intentions and behaviors. The theory planned behavior structure provides a basis for interpreting system decisions based on attitudes, social norms, and users’ perceived control over the technology used. In practical conditions, the results of this study can be implemented in a simulation-based training system for pilots, which aims to identify optimal interaction patterns with flight tracker technology