Self-learning Buildings: Leveraging Artificial Intelligence and IoT for Adaptive Architecture and Sustainable Urban Environments

Self-learning buildings, powered by artificial intelligence (AI) and Internet of Things (IoT) technologies, represent a transformative step in sustainable architecture. This study investigates the application of AI-driven models—specifically Conditional Generative Adversarial Networks (cGANs) and Diffusion Models—to enhance building operations by autonomously optimizing energy consumption, occupant comfort, and system adaptability. Through a combination of data collection from IoT sensors and advanced machine learning algorithms, the study evaluates how these models adapt building systems in real-time based on occupancy patterns and environmental conditions. The results demonstrate that the Diffusion Model significantly outperforms the cGAN Model in key areas, achieving higher energy efficiency (۰.۸۵ vs. ۰.۶۵), better occupant comfort (۰.۹۰ vs. ۰.۷۵), and improved system adaptability (۰.۹۵ vs. ۰.۸۰). These improvements are attributed to the Diffusion Model’s ability to predict future needs and make proactive adjustments to building systems. The study also explores the integration of AI models with existing Building Management Systems (BMS), revealing the practical viability of self-learning systems in real-world applications. The findings suggest that self-learning buildings offer a promising path toward reducing energy consumption, increasing occupant satisfaction, and enhancing operational efficiency. Future research could focus on multi-objective optimization, system scalability, and enhancing the transparency of AI models for broader adoption in smart cities and sustainable urban environments.

کلیدواژه ها:

Self-learning Buildings ، Artificial Intelligence (AI) ، Internet of Things (IoT) ، Energy Efficiency ، Smart Buildings