This document summarizes a study that used an artificial neural network containing two hidden layers to predict the carbonation depth of concrete. The neural network model was trained on a database of 300 concrete samples that varied in binder content, fly ash percentage, water-to-binder ratio, carbon dioxide concentration, relative humidity, square root of time, and measured carbonation depth. The results showed that the neural network performed well at predicting carbonation depth, with an R2 value of 0.9536, RMSE of 1.7761, and MAE of 1.0325. The study concluded that neural networks are a useful machine learning technique for predicting carbonation depth and could help reduce the need for physical experiments.