Automated methodological approach using artificial intelligence and fuzzy inference for qualitative sustainable and economic assessment of buildings.

Amidst the global imperative for environmental preservation and economic viability in the building sector, the integration of Life Cycle Assessment (LCA) and Life Cycle Cost Assessment (LCCA) emerges as paramount for evaluating sustainability in construction practices. This study presents an automated artificial intelligence (AI)-based framework for the environmental and economic assessment of low- to mid-rise reinforced concrete (RC) structures. The framework predicts embodied energy, carbon emissions, and life cycle cost from material quantities and basic construction information, combining quantitative and qualitative assessments. A large dataset was generated to train several regression-based machine learning (ML) models, among which XGBoost achieved the highest predictive accuracy (R² ≈ 0.99) for LCA and LCCA indicators. To enhance interpretability, classification models and a fuzzy inference system (FIS) translated the ML predictions into user-friendly sustainability-economy classes (A-C) using predefined thresholds. Validation on benchmark case studies of three, six, and ten-story RC buildings demonstrated consistent accuracy above 90% in both numerical predictions and qualitative classifications. These results confirm the robustness and reliability of the framework, offering a computationally efficient and scalable tool for preliminary sustainability and cost evaluations of RC buildings, with potential for extension to other materials and structural systems in future work.