The Summer School on Edge Artificial Intelligence is a four-day, in-person event taking place from 1st to 4th September 2025 at KTH Royal Institute of Technology in Stockholm, Sweden. It is organized under the Nordic University Cooperation on Edge Intelligence and is co-sponsored by TECoSA and Digital Futures. The program brings together leading experts and researchers in the field of Edge AI to deliver keynotes, technical talks, and hands-on exercise sessions. Each day is structured to begin with a keynote address, followed by a technical talk, and an afternoon practical exercise session—providing a blend of theoretical insights and real-world application. Topics covered include tinyML (machine learning on resource-constrained devices), intelligent offloading and inference, federated learning, and explainable and trustworthy AI at the edge—highlighting the latest advances and challenges in bringing AI capabilities closer to the data.

The event features prominent speakers from globally recognized institutions as well as industry such as V.J. Reddi from Harvard University, Eyal De Lara from University of Toronto, Henning Schulzrinne from Columbia University, Rafia Inam from Ericsson, among others. Participants will have the opportunity to engage directly with experts during Q&A discussions and deepen their understanding through interactive exercises. A social event is scheduled on the third evening to foster networking and collaboration. Designed for graduate students, PhD candidates, early career researchers, and professionals interested in distributed intelligence and real-time decision-making on the edge, this summer school offers an excellent platform to explore the future of AI beyond the cloud.

Keynote 1 : Eyal De Lara, University of Toronto, Canada

Title: Data Processing on the Edge

Short Bio: Eyal de Lara is a Professor in the Department of Computer Science at the University of Toronto. His focus is on experimental research on mobile and pervasive computing systems. Prof. de Lara served as editor in chief of ACM GetMobile the flagship publication of ACM SigMobile. His research has been recognized with an IBM Faculty Award, an NSERC Discovery Accelerator Award, the 2012 CACS/AIC Outstanding Young Computer Science Researcher Prize, and two best paper awards.

Talk 1: Jaya Prakash Champati, University of Victoria, Canada

Title: Toward Reliable and Efficient Inference in Edge AI Systems

Abstract: In the past decade, Deep Learning (DL) has achieved unprecedented improvement in the inference accuracy for several hard-to-tackle applications such as natural language processing, image classification, object detection, and identification. The state-of-the-art DL models that achieve close to 100% inference accuracy are large requiring gigabytes of memory to load them. On the other end of the spectrum, the tinyML community is pushing the limits of compressing DL models to embed them on memory-limited IoT devices. Performing inference locally on the devices reduces delay, saves network bandwidth, and improves the energy efficiency of the system, but it suffers in terms of low QoE as the small-sized DL models have low inference accuracy. To reap the benefits of doing local inference while not compromising on the inference accuracy, we explore the idea of Hierarchical Inference (HI), wherein the local inference is accepted only when it is correct, otherwise, the data sample is offloaded. However, it is generally impossible to know if the local inference is correct or not a priori. In this talk, for the prototypical image classification application, I will present the HI online learning framework for identifying incorrect local inferences. The resulting problem turns out to be a novel partitioning experts problem with continuous action space. I will present algorithms with sub-linear regret analysis and use simulation to demonstrate the efficacy of HI on ImageNet and CIFAR-10 datasets.

Short Bio: Jaya Prakash Champati is an assistant professor at the Department of Computer Science, University of Victoria (UVic), Canada. Before his current position, he was a research assistant professor at IMDEA Networks Institute in Spain. He conducted his postdoctoral research at the EECS department of KTH Royal Institute of Technology in Sweden, where he made significant contributions to the field of Age of Information Analysis and Optimization. He earned his PhD in ECE from the University of Toronto, Canada, in 2017, where his research on scheduling parallel processors earned him the Doctoral Completion Award and the Paul

Biringer Scholarship. Dr. Champati holds a master’s degree in technology from the Indian Institute of Technology (IIT) Bombay, India. He also has industry experience at Broadcom Communications, where he played a key role in developing the 4G LTE MAC layer. He was a Marie Skłodowska-Curie Actions (MSCA) postdoctoral fellow and received the Best Paper Award at the IEEE National Conference on Communications in India in 2011. His current research focus is on efficient edge-centric AI.

Keynote 2: Henning Schulzrinne, Columbia University, New York, USA

Title: Edge computing – always just beyond the horizon?

Abstract: Edge computing has attracted both academic and industrial attention for more than 20 years. For example, there are about 19,000 references to edge computing in IEEE Xplore. However, deployment has been less impressive, beyond specialized CDNs. In particular, dreams by traditional communication service providers to tap into the cloud computing revenue streams have, so far, not materialized. This talk explores some of the challenges faced by edge computing in practice: The economics are challenging, particularly as some enterprises are discovering that on-premises computing is more cost-effective than the public cloud. Many applications that need very low latency also need extremely high reliability and predictable access, encouraging fate-shared applications. Finally, the latency advantage of edge computing keeps shrinking as the hyperscalers distribute traditional data centers more geographically as power demands encourage geographic diversification, putting much of the population within 20 ms RTT of traditional cloud infrastructure. I will discuss whether edge computing can find its ecological niche, which services are most likely to benefit and how such services need to be structured to attract applications and developers.

Short Bio: Prof. Henning Schulzrinne, Levi Professor of Computer Science at Columbia University, received his Ph.D. from the University of Massachusetts in Amherst. He worked at AT&T Bell Laboratories and GMD-Fokus (Berlin), before joining the Computer Science and Electrical Engineering departments at Columbia University. He served as chair of the Department of Computer Science and as Engineering Fellow, Technology Advisor and Chief Technology Officer at the US Federal Communications Commission (FCC) from 2010 to 2017. In 2019-2020, he worked as a Technology Fellow in the US Senate and from 2022-2024 at NTIA as a Broadband Advisor.

He has published more than 250 journal and conference papers, and more than 70 Internet RFCs. Protocols co-developed by him, such as RTP, RTSP and SIP, are used by almost all Internet telephony and multimedia applications.

He is a Fellow of the ACM and IEEE, has received the New York City Mayor’s Award for Excellence in Science and Technology, the VON Pioneer Award, TCCC service award, IEEE Internet Award, IEEE Region 1 William Terry Award for Lifetime Distinguished Service to IEEE, the UMass Computer Science Outstanding Alumni recognition, is a member of the Internet Hall of Fame, has received an honorary doctorate from the University of Oulu and the ACM SIGCOMM Award.

Talk 2: Xiang Su, University of Helsinki, Finland

Title: Addressing System and Model Heterogeneity in Federated Learning

Short Bio: Dr. Xiang Su is an Associate professor in University of Helsinki, Finland. He was an associate professor at Norwegian University of Science and Technology, Norway. Dr. Su has published over 100 papers in prestigious venues, garnering more than 2,900 citations. His research expertise spans edge intelligence, mobile computing, extended reality, and Internet of Things.

Keynote 3: Vijay Janappa Reddy, Harvard University, USA

Title: Generative AI at the Edge

Short Bio: Vijay Janapa Reddi is an Associate Professor in the John A. Paulson School of Engineering and Applied Sciences at Harvard University. Prior to joining Harvard University, he was an Associate Professor at The University of Texas at Austin. His research interests include computer architecture and runtime systems, specifically in the context of edge and mobile computing systems (smartphones, autonomous vehicles, aerial robots, etc.) to improve their performance, power efficiency, and reliability. Dr. Janapa Reddi is a recipient of multiple honors and technical achievement awards, including the MICRO and HPCA Hall of Fame (2018 and 2019, respectively), the National Academy of Engineering (NAE) Gilbreth Lecturship Honor (2016), IEEE TCCA Young Computer Architect Award (2016), Intel Early Career Award (2013), Google Faculty Research Awards (2012, 2013, 2015, 2017), Best Paper at the 2005 International Symposium on Microarchitecture (MICRO), Best Paper at the 2009 International Symposium on High Performance Computer Architecture (HPCA), and IEEE’s Top Picks in Computer Architecture awards (2006, 2010, 2011, 2016, 2017). Beyond his technical research contributions, Dr. Janapa Reddi is passionate about STEM education at early age. He is responsible for the Austin Independent School District’s “hands-on” computer science (HaCS) program, which teaches 6th- and 7th-grade students programming and the high-level principles governing a computing system using open-source prototyping platforms like Arudinos. He received a BS in computer engineering from Santa Clara University, an MS in electrical and computer engineering from the University of Colorado at Boulder, and a Ph.D. in computer science from Harvard University.

Talk 3: Roberto Morabito, Eurecom, France

Title: From Edge to Tiny: Reimagining AI in the Era of Generative and Embedded Intelligence.

Short Bio: Roberto Morabito is an Assistant Professor in the Communication Systems Department at EURECOM, France. His research focuses on networked systems, edge computing, and distributed AI, with recent work exploring the role of generative AI in constrained environments. He holds a Ph.D. in Networking Technology from Aalto University and has held positions at Ericsson Research, the University of Helsinki, and Princeton University. He has also been a visiting researcher at INRIA, TUM, and Yale University.

Keynote 4: Rafia Inam, KTH/Ericsson, Sweden

Title: The role of Trustworthy AI / Explainable AI in Telecom industry

Abstract: Trust and reliance on modern telecom systems are widespread. However, the adoption of AI introduces new risks and necessitates countermeasures. Governments, companies, and regulatory bodies worldwide are recognizing the need for trustworthy AI systems.

The presentation will discuss the importance of AI regulation and its impacts, and current standardization work in Europe. It will also include Trustworthy AI and Explainable AI for Telecom industry to enable customer trust; and how these techniques can support the industry to ensure correctness of AI models, provide transparency to different users, enable automation of telecom use cases, and help to identify and describe unexplained or new behavior of the models.

Short Bio: Rafia Inam is a Senior Research Manager at Ericsson Research in Trustworthy AI and Adjunct Professor at The Royal Institute of Technology (KTH), Sweden. She has conducted research for Ericsson for the past ten years on 5G for industries, network slices, and network management; and AI for automation. She specializes in trustworthy AI, Explainable AI, risk assessment and mitigations using AI methods, and safety for cyber-physical systems for telecom and CPS. She is also contributing to trustworthy AI based standardization specially to European standards in CEN/CLC based on EU AI Act.

Rafia received her PhD in predictable real-time embedded software from Mälardalen University in 2014. She has co-authored 55+ refereed scientific publications and 60+ patent families, and 2 best paper awards. She won Ericsson Top Performance Competition 2021 on her work on AI for 5G network slice assurance and was awarded multiple Ericsson Key contributor awards.