Title:

Transforming edge computing systems to intelligent systems using TinyML

Keynote speaker:

Dr. Athanasios Kakarountas, University of Thessaly, Greece

Athanasios (Thanos) Kakarountas is an Associate Professor holding a position for “Embedded Computing Systems” at the Department of Computer Science and Biomedical Informatics, University of Thessaly. He is the Head of the Department and Director of the Intelligent Systems Laboratory and the Vice Chair of the Research and Innovation Regional Committee for the Region of Central Greece. His research interests include green edge Computing, Reconfigurable Computing, Embedded Systems design for IoT, Health, Smart Cities, and Security applications. He is a co-chair of the eGOV Special Technical Committee of the IEEE Computer Society. In the past, he served as the Deputy Chairman of the National Infrastructures for Research and Technology, the Chairman of the Research and Innovation Regional Committee for the Region of Central Greece, and the Vice-Chair of the Board of the Greece IEEE Section. He was also the interim Chairman of the Consumer Technology Society – Greece Chapter. He is Editor in Chief at the IEEE Potentials magazine. He has participated in more than 30 research projects. He has published over 150 articles in international journals, conference proceedings, and seven technical book chapters.

Abstract:

Edge AI is providing the framework and the means to integrate Machine Learning (ML) models to devices of restricted resources. Without the need for cloud computing infrastructures, the high dependability of telecommunication equipment and the excessive processing power required by today’s intelligent systems, Edge AI is promising a tailored solution to many applications. Among the latter, safety critical applications can benefit by the characteristics of the developed technology for Edge AI. This talk focuses on the emerging technology of TinyML, which allows easy integration of ML models to embedded systems, offering higher autonomy, data integrity and security, and low latency to derive a result.