EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Environmentally sustainable framework integrating agentic intelligence and federated learning for occupancy-driven smart building energy management

Prabhu Rajaram, Gnana Swathika Odiyur Vathanam

International Journal of Low-Carbon Technologies · 27.01.2026 · 21

DOI: 10.1093/ijlct/ctag159 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Low-Carbon Technologies, online erschienen am 12.11.2026. Verfasst von Prabhu Rajaram und 1 weiteren Autor:innen. Die Arbeit nennt 34 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Sustainable smart city infrastructure requires efficient and intelligent energy management strategies for smart buildings. This study presents a unified Sustainability-Aware Agentic Federated Learning framework designed to improve occupancy-driven energy management while supporting sustainability objectives. Unlike conventional methods that often evaluate artificial intelligence techniques separately, the proposed framework combines federated learning (FL) and agentic decision-making within a single architecture to support privacy-aware and adaptive building operations. The framework introduces a multi-objective FL strategy that not only emphasizes predictive accuracy but also delves into energy consumption and carbon emissions within the optimization process. In addition, an adaptive aggregation mechanism is developed to dynamically adjust the contribution of participating clients based on their predictive performance, energy efficiency, and data quality. To further enhance operational intelligence, an agentic reinforcement learning layer is integrated to enable autonomous heating, ventilation, and air conditioning, and building control decisions through continuous interaction with the environment. The proposed framework is evaluated using a standard occupancy detection dataset and compared with several benchmark approaches, including classical machine learning, deep learning, FL, and hybrid models under the same experimental conditions. The results demonstrate that the proposed approach achieves reliable occupancy prediction while improving energy efficiency and reducing carbon emissions. In particular, the framework achieves effective energy savings of more than 84%, highlighting its potential for sustainable and intelligent smart building energy management. The proposed work also supports the objectives of Sustainable Development Goals (SDGs), particularly SDG 7 (Affordable and Clean Energy), SDG 11 (Sustainable Cities and Communities), and SDG 13 (Climate Action).

Zitieren

APA

Rajaram, P., & Vathanam, G. S. O. (2026). Environmentally sustainable framework integrating agentic intelligence and federated learning for occupancy-driven smart building energy management. International Journal of Low-Carbon Technologies, 21. https://doi.org/10.1093/ijlct/ctag159

Vancouver

Rajaram P, Vathanam GSO. Environmentally sustainable framework integrating agentic intelligence and federated learning for occupancy-driven smart building energy management. International Journal of Low-Carbon Technologies. 2026;21. doi:10.1093/ijlct/ctag159

BibTeX
@article{Rajaram2026,
  author = {Rajaram, Prabhu and Vathanam, Gnana Swathika Odiyur},
  title = {Environmentally sustainable framework integrating agentic intelligence and federated learning for occupancy-driven smart building energy management},
  journal = {International Journal of Low-Carbon Technologies},
  year = {2026},
  volume = {21},
  doi = {10.1093/ijlct/ctag159}
}
Kostenlos starten →