This research proposes a federated learning system that prioritizes participation from solar-powered edge nodes. It reduces training overhead and promotes low-carbon distributed intelligence for climate-conscious data processing in modern edge networks.

Ramesh Desai, Jean Claude Laurent, Priya Malvika Nair, Carlos Fernando Gutierrez, Fatima Noor Al-Hakim

We introduce a thermal-aware scheduler that learns heat profiles and redistributes AI tasks to avoid thermal throttling. It enhances energy efficiency and device longevity, enabling sustainable real-time AI processing in mobile environments.

Parthiban Nair, Fatima Noor Al-Kareem, Jean Michel Laurent, Priya Sangeetha Ramanathan, Carlos Antonio Delgado

This research presents an AI-driven task placement system. It uses real-time emission analytics and workload profiling to dynamically shift compute loads, reducing carbon output across distributed cloud infrastructures.

Olivia Sanders, James Holloway, Rebecca Dalton, Thomas Finch, William Cooper

We propose a battery-free monitoring platform for forests. Using solar-harvesting IoT and compact AI models, it identifies ecological anomalies and ensures sustained, maintenance-free operation across large conservation zones.

Zhang Rui Liang, Gao Tian Qiang, Xu Wen Bo

This study introduces a circular AI system that forecasts component failure and suggests reuse paths. It reduces electronic waste and supports sustainable hardware design through targeted repair and intelligent part salvaging.

Jean Andre Morel, Priya Devika Subramaniam, Fatima Noor Al-Salem, Carlos Javier Ramirez

We introduce a smart traffic platform integrating solar-powered edge devices with lightweight AI. It reduces congestion, lowers vehicle emissions, and supports real-time control in urban transit networks powered by green infrastructure.

Matthew Quinn, Rachel Dawson, Brian McKinley, Susan Hopkins

We present an AI-based load prediction model for smart microgrids. It reshapes energy consumption in remote zones, optimizing renewable use and ensuring continuous, low-emission rural electrification.

Jean Claude Girard, Priya Shalini Sundaram, Fatima Noor Al-Zaid, Carlos Emmanuel Torres

This paper introduces a task scheduler that prioritizes jobs based on available renewable energy. It enables energy-efficient inference on edge devices while reducing reliance on the main grid in remote environments.

Isabelle Grant, Henry Clarkson, Megan Wallace, Owen Baxter, Daniel Forbes