A team of researchers from the University of Mississippi has created artificial intelligence (AI) algorithms capable of predicting the likelihood that potholes will develop on roadways.

To develop the algorithm, the team tested different algorithms’ abilities to forecast how well asphalt pavements made with reclaimed materials withstood moisture, which tends to weaken and subsequently destroy asphalt.

“We focused on moisture damage, which is one of the most critical issues in asphalt pavements, particularly for wet and cold regions, because it results in a variety of distresses like stripping, potholes and cracking,” explained the researchers. “We evaluated the effectiveness of four different artificial intelligence algorithms in predicting moisture damage in asphalt mixtures containing (reclaimed asphalt pavement) materials.

Based on this assessment, the team concluded that the algorithms effectively and accurately predicted moisture damage in asphalt mixtures. As such, they determined that they could use this information to optimize material selection and subsequently make predictions about failure probability in the pavement’s life cycle.

The study, "Prediction of moisture susceptibility of asphalt mixtures containing RAP materials using machine learning algorithms," appears in the International Journal of Pavement Engineering.

To contact the author of this article, email mdonlon@globalspec.com