Applications of Artificial Intelligence in the Selection and Evaluation of Building Stones for Sustainable Urban Development

Applications of Artificial Intelligence in the Selection and Evaluation of Building Stones for Sustainable Urban Development

Hengameh asadi1

1) M.Sc. Student in Civil Engineering, Department of Civil Engineering, Science and Research Branch, Islamic Azad University, Tehran, Iran

IOI Code : ABKD-ACZGGD
Publication : Fourth International Conference on applied researches in civil engineering, architecture and urban planning (icarcau.com)
Abstract :
    Rapid urbanization and the increasing demand for construction materials have heightened the importance of the scientific and precise selection of building stones, which remain one of the oldest and most widely used materials in civil and urban projects. Traditional evaluation methods, which rely heavily on time-consuming and costly laboratory tests, exhibit limited efficiency when faced with massive datasets and complex relationships between physical, mechanical, environmental, and economic properties. This research, employing a review-analytical approach, investigates the role of Artificial Intelligence (AI) technologies—including Artificial Neural Networks (ANN), Adaptive Neuro-Fuzzy Inference Systems (ANFIS), Random Forest (RF), Support Vector Machines (SVM), and Deep Learning (DL)—in predicting mechanical properties, assessing durability and reliability, and supporting decision-making for selecting building stones aligned with sustainable urban development principles. The proposed conceptual framework categorizes evaluation criteria into four domains: physical-mechanical, durability and reliability, environmental, and economic-functional, illustrating how these criteria are integrated through machine learning models and multi-criteria decision-making (MCDM) methods. Findings from the literature review indicate that ANN-based models and ensemble methods, such as Random Forest and Gradient Boosting, demonstrate significant accuracy in predicting uniaxial compressive strength, porosity, and abrasion resistance from non-destructive tests, potentially reducing laboratory costs and time. Finally, existing challenges, such as the lack of standardized databases, the difficulty of model interpretability, and the lack of integration between Geographic Information Systems (GIS), Building Information Modeling (BIM), and AI tools, are discussed, alongside suggestions for future research.

Keywords : artificial intelligence machine learning building stone sustainable urban development artificial neural networks material evaluation