Citation

Sungheetha, Akey and Rajagopal, Kayapati (2026) Energy-Aware Explainable AI Framework for Sustainable Cloud Computing: A Novel Approach to Green Machine Learning with Real-Time Carbon Footprint Optimization. In: 2026 7th International Conference on Mobile Computing and Sustainable Informatics (ICMCSI). Institute of Electrical and Electronics Engineers Inc., pp. 1860-1870. ISBN 979-833155519-1

Text IEEE Xplore Full-Text PDF_12.pdf - Published Version Restricted to Repository staff only Download (371kB)

Abstract

his research introduces a novel Energy-Aware Explainable AI (EA-XAI) framework that integrates carbon footprint optimization with interpretable machine learning for sustainable cloud computing environments. The proposed system achieves a remarkable 34.7% reduction in energy consumption while maintaining model accuracy at 94.2% and providing realtime explanations with SHAP values computed in 12.3ms average latency. The framework incorporates green computing principles through dynamic resource allocation algorithms that optimize carbon emissions (CO reduction of 2.1 tons/month), renewable energy utilization efficiency of 87.3%, and cost optimization achieving 41.6% reduction in operational expenses. The methodology addresses critical challenges in sustainable AI deployment by implementing novel carbon-aware scheduling algorithms, energy-efficient model compression techniques achieving 67.8% model size reduction, and transparent decision-making processes suitable for industrial applications requiring regulatory compliance and environmental sustainability certifications.

Item Type:	Book Section
Uncontrolled Keywords:	Explainable AI, Sustainable Computing
Subjects:	Q Science > QA Mathematics > QA71-90 Instruments and machines > QA75.5-76.95 Electronic computers. Computer science
Divisions:	Faculty of Information Science and Technology (FIST)
Depositing User:	Ms Rosnani Abd Wahab
Date Deposited:	04 May 2026 04:07
Last Modified:	04 May 2026 04:07
URII:	http://shdl.mmu.edu.my/id/eprint/15859

Downloads

Edit (login required)