This research introduces a reinforcement learning framework hosted on the cloud for dynamic urban traffic control. The model continuously learns traffic patterns, adapting signal phases in real time to optimize flow and minimize congestion in metropolitan infrastructure systems.

Ananya Pradeep Chatterjee, Lars Viktor Sundstrom, Zhen Yu Liu, Gabrielle Noelle Ferguson, Omar Abdulrahman El-Tayeb

We propose a novel cost-aware placement algorithm for Kubernetes containers in shared cloud environments. It balances resource utilization with tenant priorities, contributing to operational sustainability and minimizing cloud service expenditures through intelligent workload orchestration.

Matteo Luciano Ferraro, Sara Meera Iqbal, Xue Fang Chen, Daniel Joseph Clancy, Helena Grace Whitmore, Prakash Naresh Bhandari

This paper details an AI-driven platform that enables predictive maintenance of renewable energy systems. Leveraging cloud-based sensor analytics, it anticipates equipment failures and optimizes maintenance schedules, thereby enhancing energy reliability and reducing operational downtimes in solar and wind infrastructures.

Hiroshi Kojiro Yamazaki, Lucas Emmanuel Dupont, Rajiv Kumar Manohar, Natalie Louise Morin, William Charles Newton

We present a cloud-integrated model that enhances cybersecurity by enabling collaborative threat intelligence exchange. It detects real-time anomalies across interconnected systems, improving defensive postures and supporting proactive mitigation strategies within multicloud operational frameworks.

Jing Wei Zhou, Olivia Mae Bradshaw, Andrej Milan Novak, Fatima Zehra Qadri, Christopher Dean Holloway

A new scheduling algorithm for machine learning tasks across heterogeneous cloud clusters is introduced. It dynamically assigns workloads based on latency constraints and hardware capabilities, improving training efficiency and reducing inference lag across distributed AI applications.

Siddharth Ajay Raghavan, Emma Claire Bennet, Jiao Feng Liu, Yannick François Laurent, Ayesha Noor Hassan

This paper introduces a decentralized authentication protocol leveraging blockchain for IoT devices across diverse platforms. It ensures secure identity verification and access management while maintaining auditability and privacy in distributed edge-cloud environments handling sensitive industrial or consumer applications.

Takumi Ryota Hoshino, Isabelle Charlotte Greene, Liang Zhi Qiu, Fernando Luis Ramirez, Naomi Bridget Callahan

We present a graph neural network-based approach to forecast workload demand in edge-cloud infrastructures. Our model facilitates predictive auto-scaling of services, reducing idle consumption while meeting real-time performance requirements, enabling greener and smarter deployment in virtualized computing architectures.

Yasmin Farida Chowdhury, Gregor Hans Meier, Wei Zhong Tan, Patrick Jerome O'Donnell, Junaid Rahimullah Khan