The road network is subject to various mechanical loads which call into question its easement, its durability and the safety it provides. On this, it is essential to carry out an accurate prediction of its behavior in order to analyze and model the response of roadway structures following the application of mechanical loads. To achieve this, we used artificial neural networks and Monte Carlo simulation. By combining the two methods, our study presents a comprehensive approach to predict the behavior of roadways. This combination resulted in very accurate and reliable predictions. Specifically, the model demonstrated its effectiveness with an exceptionally low mean square error and a perfect coefficient of determination value of 1, indicating accurate and reliable results. Thus, the comparison of the established model with other models using cross-validation highlighted the accuracy of the prediction of the established model. The distribution generated by the simulation gave us a clear idea about the consistency and reliability of the chosen material. This analysis, which is part of scientific research in the field of infrastructure, is a research avenue to ensure the longevity of roadway structures.