Application of artificial intelligence in HVAC system optimization: opportunities, challenges, and perspectives
DOI:
https://doi.org/10.3846/da.2025.029Keywords:
artificial intelligence, energy savings, fault detection and diagnostics, heating, optimization of engineering systems, thermal comfort maintenance, ventilationAbstract
Artificial Intelligence (AI) technologies are rapidly advancing, and their application in optimizing the heating, ventilation, and air conditioning (HVAC) systems operation is becoming more and more common. HVAC systems are using a significant portion of a building‘s energy consumption, moreover systems failures or poor management can lead to substantial energy losses. AI technologies offer a potential to ensure efficient energy use, improve user thermal comfort parameters, and contribute to sustainability efforts. By utilizing real-time data analysis and forecasting methods, AI can optimize HVAC system performance, reduce energy usage, and enhance system reliability, also extend their operational lifespan, by effectively predicting failures, enabling preventive measures and avoiding costly repair expenses. This paper presents a review of the latest literature, examining key AI models and their applications to HVAC systems. Despite the progress in these technologies, multiple challenges still remain, such as model accuracy, adaptability, integration with existing systems, lack of data or it’s quality, model interpretation, high implementation costs, and ensuring user comfort. Nevertheless, AI undoubtedly opens up a lot of new opportunities and become an inevitable part of the future of HVAC systems. Although certain obstacles are currently encountered, technological advancements and the continuous implementation of new solutions will enable even greater application of AI. Therefore, its integration into HVAC systems is not only a technological advancement but also a significant step toward a more sustainable, efficient, and comfortable environment in the future.
Downloads
Published
Conference Event
Section
Copyright
License

This work is licensed under a Creative Commons Attribution 4.0 International License.