International Journal of Sustainable Building Technology and Urban Development ISSN:2093-761X(Print) 2093-7628(Online)

The growing need for sustainable and energy-efficient built environments requires intelligent systems for real-time monitoring and optimization of building performance. This study presents an IoT sensor-based artificial intelligence (AI) framework for sustainable building performance optimization using Random Forest prediction modeling for crop recommendation and XGBoost for fertilizer suggestion. The proposed framework integrates data from IoT-enabled sensors, including temperature, humidity, air quality, occupancy, and energy consumption, to enable data-driven decision-making in smart buildings. Machine learning models such as Decision Tree, Random Forest, and XGBoost are employed to predict optimal operational conditions for enhancing environmental performance and resource efficiency. The datasets, derived from publicly available sources and simulated building environments, ensure accessibility and reproducibility. Model performance is evaluated using standard metrics, including accuracy, precision, recall, and F1-score. Among the models, Random Forest achieves the highest accuracy of 99.55%, outperforming Decision Tree and XGBoost for crop suggestion, and XGBoost achieves a higher accuracy 99.46% for fertilizer suggestion than Random Forest and Decision Tree. The proposed approach supports sustainable building management by improving energy efficiency, indoor environmental quality, and operational decision-making. It aligns with key aspects of environmental performance assessment, facility management, and smart urban development. Future work may incorporate deep learning and hybrid models to enhance scalability across diverse building types and climatic conditions.