Carbon and Cooling Efficiency Scheduler for Sustainable Cloud Operations

DOI

10.2991/978-94-6239-616-6_13How to use a DOI?

Keywords

Carbon-Aware Scheduling; LSTM Forecasting; ElectricityMaps SDK; Cloud Sustainability; Streamlit UI; Ensemble Learning

Abstract

Data centers worldwide use amounts of electricity that contributes to ever-increasing carbon emissions that change by time and across regions. Currently, cloud computation schedulers do not account for the real-time carbon intensity associated with electricity usage. This regional variability leads to an over-scheduling of workloads during periods of high overall and regional carbon intensity, causing excess environmental damage. We explore a novel hybrid carbon-aware scheduling framework that uses machine learning in combination with regular carbon data, to intelligently schedule cloud workloads. This architecture relies on the historical forecasting capability of a locally trained LSTM model with a real-time carbon intensity forecast available through the ElectricityMaps SDK. A weighted ensemble method for both the local trained forecasting model and real-time permit forecasting update of the carbon emission forecast is applied to work on real-time carbon emissions forecasts, and greatly improves the service offered through the work of additional forecasting solutions work. Overall, the framework describes a work solution for scheduling and scheduling workloads in a cloud agnostic manner for single and multi-regional decisions in order to achieve period averages of reduced overall carbon emissions, including average carbon for resources wether repair to carbon intensity. By coupling real-time predictions via the SDK with predictive forecasting, flow in this innovative modular scheduling pipeline, and modeling the real-time carbon emissions deeply into this workflow environment, user engagement and training interface provides improved predictive accuracy of carbon emissions with the assigned unloaded workloads and reduced overall emissions through more thoughtful planning, providing an innovative action toward a faster, real-timely, solution journey toward sustainable collapse computing.

Copyright

© 2026 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - G. Prabu
AU  - B. Mona
AU  - V. Nandini
AU  - M. Smitha Keren
PY  - 2026
DA  - 2026/03/31
TI  - Carbon and Cooling Efficiency Scheduler for Sustainable Cloud Operations
BT  - Proceedings of the International Conference on Artificial Intelligence and Secure Data Analytics (ICAISDA 2025)
PB  - Atlantis Press
SP  - 152
EP  - 164
SN  - 1951-6851
UR  - https://doi.org/10.2991/978-94-6239-616-6_13
DO  - 10.2991/978-94-6239-616-6_13
ID  - Prabu2026
ER  -