Speaker: Dr. Öğr. Üyesi Ümit Demirbaga

Abstract: The rapid convergence of artificial intelligence, the Internet of Things, and cloud computing is accelerating the development of Industry 5.0 applications, while simultaneously intensifying the energy and carbon footprint of large-scale computing infrastructures. This talk presents EcoCloud, a green computing framework designed to improve energy and carbon efficiency in Industrial IoT-enabled cloud environments. The proposed approach combines real-time resource monitoring, a multilayer perceptron (MLP)-based energy prediction model, and an ant colony optimisation (ACO)-based task scheduling strategy to dynamically schedule MapReduce workloads on cloud-based Hadoop clusters.
The seminar will discuss the architectural design of EcoCloud, the underlying energy model, and the integration of predictive intelligence into task scheduling decisions. Experimental evaluations demonstrate that the proposed framework can reduce energy consumption and improve execution efficiency when compared with traditional scheduling methods. In particular, the results indicate approximately 25% energy reduction and 20% improvement in makespan, highlighting the potential of intelligent scheduling for sustainable cloud operations.
Overall, the talk will emphasise how AI-driven optimisation can support greener Industrial IoT ecosystems and contribute to environmentally responsible cloud computing.

Bio: Ümit Demirbaga is an Assistant Professor of Computer Engineering at Bartin University, Türkiye. Before this, he worked as a Postdoctoral Research Associate in Health Data Science in the Department of Medicine at the University of Cambridge, UK, fully funded and employed by the university from 2022 to 2024. He was also a Visiting Postdoctoral Fellow at the European Bioinformatics Institute (EMBL- EBI), UK, concurrent with his tenure at the University of Cambridge. He received his BSc in Electronic and Computer Education Department from Marmara University, Türkiye, in 2011 and his MSc and PhD in Computer Science from Newcastle University, UK, in 2017 and 2022, respectively. His research interests are big data analytics, cloud computing, distributed systems, machine learning, eXplainable AI, generative AI, digital twins, data science, and energy-efficient computing systems. Dr Demirbaga was awarded the Colin Graham Outstanding Performance Award with Best Team Project Awar d in his MSc in 2017 and received the 2022 IEEE TCSC Outstanding PhD Dissertation Award in Scalable Computing. Additionally, he won first place at the TEKNOFEST Artificial Intelligence in Health Competition, University and Above-Level, Bioinformatics Analysis Development Category, Türkiye.

https://us06web.zoom.us/j/84636366540
MEETING ID: 84636366540
PASSCODE: 25862