CINTECX hosted on April 29 the doctoral thesis defense of Ahmed Mossad Ibrahim Elseicy, a researcher in the field of civil engineering, who presented a work focused on the use of deep learning techniques to improve the analysis of pavement condition. The defense took place in the center’s auditorium, within the PhD program Applied Geotechnologies for Construction, Energy and Industry.
Under the title Deep Learning-based interpretation of ground penetrating radar for pavement condition monitoring, the research integrates infrastructure engineering and artificial intelligence. The work is based on the use of ground-penetrating radar (GPR), a non-destructive technique commonly used to analyze the interior of pavements, and addresses one of its main challenges: the complexity and time required to correctly interpret the data obtained.
The thesis, supervised by Henrique Lorenzo Cimadevilla and Mercedes Solla Carracelas, proposes new methods to automate this interpretation process, reducing reliance on manual analysis and increasing the reliability of the results. To achieve this, Elseicy trains deep learning models capable of recognizing patterns, identifying structural layers, and detecting deterioration that is invisible to the naked eye.
From radar to digital models of infrastructures
Just as a person learns to identify objects in images, the models developed by Elseicy learn to differentiate pavement layers, detect internal damage, or locate anomalies that cannot be seen with the naked eye.
One of the most notable advances is the ability to transform this data into digital models, pavement condition maps, and three-dimensional representations that make interpretation easier, providing a clear view of the infrastructure’s actual condition.
The advantage of these tools lies in the wide range of direct applications they offer for both roads and bridges, enabling the detection of early deterioration or hidden structural elements, such as reinforcements, without the need for physical intervention on the structure. Elseicy’s research improves maintenance planning, optimizes resources, and increases safety by anticipating structural problems and facilitating decision-making regarding infrastructure upkeep.