Advancing Concrete Segregation Detection: Leveraging Explainable AI (XAI) for Improved Accuracy and Transparency

In the aftermath of the devastating seismic event that struck Morocco in September 2023, resulting in the collapse of numerous fragile structures and a significant loss of human life, the urgency to strengthen disaster preparedness and response capabilities has never been more critical. This juncture provides a unique opportunity to leverage the transformative potential of cutting-edge technologies, particularly Artificial Intelligence (AI) with a specific emphasis on Explainable AI (XAI) methods. Concrete, being an essential construction material valued for its notable strength, durability, and versatility, sees extensive use in various infrastructure projects. However, the presence of concrete segregation presents a substantial challenge in both the construction and upkeep of concrete structures. Concrete segregation entails the uneven distribution of aggregates within the concrete matrix, which can lead to adverse effects. It jeopardizes structural integrity, diminishes performance, and potentially poses safety hazards. Given these significant concerns, it is imperative to identify and quantify instances of concrete segregation. In response to this requirement, the current investigation introduces a novel method for assessing concrete segregation using an image processing-based technique, incorporating Deep Learning and Explainable Artificial Intelligence. This research not only improves the precision and reliability of detecting concrete segregation but also ensures transparency in the decision-making process. This innovative approach has the potential to transform concrete construction and maintenance practices, ensuring the structural integrity and performance of infrastructure projects are upheld while minimizing safety risks linked to concrete segregation.