MobiUK 2025 Conference Programme

Monday 7th July

Event Location: G.07, Informatics Forum, University of Edinburgh

9:30 – 10:10

Registration and Coffee

10:10 – 10:20

Welcome and Opening Remarks (Mahesh Marina)

10:20 – 11:00

Invited Talk by Silvia Del Din, Newcastle University (Chair: Cecilia Mascolo)

Measuring mobility outside the laboratory with digital health technology: are we there yet?

 

Have you ever thought about how you walk? Did you know that mobility and the way you walk – (your gait) is considered the 6th vital sign, acting as a biomarker for health status and brain function? Mobility is emerging as a powerful tool to detect early risk and monitor disease progression across a number of diseases (e.g., Parkinson’s disease). Typically, quantitative gait assessment has been limited to specialised laboratory facilities. However, measuring gait in home and community settings (i.e., in the “real world”) may provide a more accurate reflection of gait performance as it allows walking activity to be captured over time in habitual contexts. In this context, using digital health technology (e.g., wearable devices including Inertial Measurement Units) allows objective measurement of real-world walking activity/behaviour as well as discrete gait characteristics (e.g., gait speed). This talk will give an overview of how and why we should monitor mobility in the real-world with digital health technology, and the types of data/information can be collected remotely. Advantages and challenges of using digital health technology for remote monitoring of mobility will be discussed, including results from clinical research applications in Parkinson’s disease.

11:00 – 12:00

Session 1: Earables and Wearables (Chair: Jagmohan Chauhan)

Monitoring Walking Gait Parameters with Earables 

Jake Stuchbury-Wass (University of Cambridge), Yang Liu (University of Cambridge), Kayla-Jade Butkow (University of Cambridge), Josh Carter (University of Bath), Qiang Yang (University of Cambridge), Mathias Ciliberto (University of Cambridge), Ezio Preatoni (University of Bath), Dong Ma (Singapore Management University) and Cecilia Mascolo (University of Cambridge) 

Listening to Your Lungs: Continuous Respiration Volume Monitoring Using In-Ear Audio 

Yang Liu (University of Cambridge), Qiang Yang (University of Cambridge), Kayla-Jade Butkow (University of Cambridge), Jake Stuchbury-Wass (University of Cambridge), Dong Ma (Singapore Management University) and Cecilia Mascolo (University of Cambridge) 

A Wearable ECG Pipeline for Free-Living VO2max Prediction and HRV Interpretation 

Aron B. Syversen, Zhi-Qiang Zhang, David Jayne and David Wong (University of Leeds) 

Benchmarking Foundation Models on Out-of-Distribution Wearable Biosignals 

Andres Alvarez Olmo (University of Cambridge), Sotirios Vavaroutas (University of Cambridge), Yu Wu (University of Cambridge) and Cecilia Mascolo (University of Cambridge) 

12:00 – 13:00

Lunch and Poster Setup

13:00 – 13:40

Invited Talk by Remy Pottier, Arm (Chair: Alastair Beresford)

How the Combination of Immersive Experiences and AI Will Spark a New Revolution in Reality

The convergence of AI and immersive technologies—spanning AR, VR, MR, digital twins, and haptics—is ushering in a new era of interaction across work, play, and daily life. This talk explores the evolution from static digital interfaces to rich, real-time multimodal experiences. We’ll examine the breakthroughs in neural rendering (e.g., NeRF, Gaussian Splatting), compute-efficient pipelines, and foundational hardware enablers needed to support immersive AI applications at scale.

13:40 – 14:10

Session 2: From Devices to Networks (Chair: Veljko Pejovic)

MakeDevice: from prototype to manufacture of mobile devices 

Aron Eggens (Lancaster University), James Hahn (Lancaster University), Kobi Hartley (Lancaster University), Steve Hodges (Lancaster University) and Joe Finney (Lancaster University) 

On Deploying a Campus Scale Private 5G Open RAN Testbed 

Andrew E. Ferguson (University of Edinburgh), Ujjwal Pawar (University of Edinburgh), Tianxin Wang (University of Edinburgh) and Mahesh K. Marina (University of Edinburgh) 

14:10 – 14:40

Coffee Break

14:40 – 15:20

Invited Talk by Adrian Friday, Lancaster University (Chair: Fahim Kawsar)

Rethinking resource efficiency for sustainability

In this talk, I'll attempt to reconsider the role of 'resource efficiency' not as a driver of growth of ICT's impacts on the planet, but rather to work within planetary limits. Digital technology unquestionably transforms our understanding of the world and helps society reconfigure itself to do new things from better healthcare, climate science, self-driving cars and smart cities, and more.  Yet, I will argue, ICT has significant direct impacts on energy material systems, and an exponentially growing footprint.  As

researchers, innovators and practitioners, what should the MobiUK

community be doing to help bring about a more environmentally

sustainable future?

15:20 – 15:50

Session 3: HCI (Chair: Sarah Clinch)

Towards Capturing the Therapeutic Effects of Music in Everyday Life: Building The BEATS Dataset 

Benjamin Gutierrez Serafin (University of Glasgow), Tanaya Guha (University of Glasgow) and Fahim Kawsar (University of Glasgow & Nokia Bell Labs) 

CASES: A Cognition-Aware Smart Eyewear System for Understanding Reading 

Yingying Zhao (University of Strathclyde) and Mingzhi Dong (University of Bath) 

15:50 – 16:20

Lightning Talks for Posters

16:20 – 17:20

Poster Session and Drinks Reception in MF2 on 4th Floor, Informatics Forum (Chair: Jingjie Li)

Assessing seasonal and weather effects on depression and physical activity using mobile health data 

Yuezhou Zhang (Institute of Psychiatry, Psychology, and Neuroscience, King’s College London), Amos Folarin (Institute of Psychiatry, Psychology, and Neuroscience, King’s College London) and Richard Dobson (Institute of Psychiatry, Psychology, and Neuroscience, King’s College London) 

LungLDM: Prompt-Based Synthetic Lung Sound Generation Using Latent Diffusion Models for Respiratory Health Diagnostic 

Mohammed Mosuily (University of Southampton) and Jagmohan Chauhan (University College London) 

Developing a Mobile App to Self-Manage Postural Orthostatic Tachycardia Syndrome 

Rosalind Thwaites (University of Southampton) and Adriana Wilde (University of Southampton) 

Identifying Folk Methods Employed with Fitness Applications by Users with Chronic Mobility Issues to Improve Application Design 

Myfanwy Fflur (University of Manchester), Markel Vigo (University of Manchester) and Sarah Clinch (University of Manchester) 

ActiNet: Improved Activity Recognition in Wrist-Worn Accelerometers 

Aidan Acquah (University of Oxford), Shing Chan (University of Oxford) and Aiden Doherty (University of Oxford) 

Bayesian based Long-Tailed Continual Learning 

Hao Dai (University of Southampton), Chong Tang (University of Southampton) and Jagmohan Chauhan (University College London) 

BoTTA: Benchmarking on-device Test Time Adaptation 

Michal Danilowski (University of Birmingham), Soumyajit Chatterjee (Nokia Bell Labs) and Abhirup Ghosh (University of Birmingham) 

Setting Up a MANET with Raspberry Pi 

Tomasz Grabowski (Heriot-Watt University), Jamie Lewis (Heriot-Watt University) and Idris Ibrahim (Heriot-Watt University) 

Adding a new dimension to Large Language Scaling Laws and its implications for the edge

Preslav Aleksandrov (University of Cambridge), Meghdad Kurmanji (University of Cambridge) and Nicholas Lane (University of Cambridge) 

Federated LLM Training with Heterogeneous Mobile Clients 

Andrzej Szablewski (University of Cambridge), Lorenzo Sani (University of Cambridge) and Nicholas D. Lane (University of Cambridge) 

Resource-Efficient Knowledge Editing for Mobile LLMs 

Zhenyan Lu (Beijing University of Posts and Telecommunications), Dongqi Cai (Beijing University of Posts and Telecommunications), Chen Peng (Beijing University of Posts and Telecommunications), Zexi Li (University of Cambridge), Shanggua Wang (Beijing University of Posts and Telecommunications), Nicholas D. Lane (University of Cambridge) and Mengwei Xu (Beijing University of Posts and Telecommunications) 

Reimagining Databox with User-Facing Agents 

Josh Millar (Imperial College London), Amir Al Sadi (Imperial College London), Hamed Haddadi (Imperial College London), Ryan Gibb (University of Cambridge) and Anil Madhavapeddy (University of Cambridge) 

Exploring Privacy and Security Perceptions and Practices of Migrant Domestic Workers in Multi-User Smart Homes 

Shijing He (King’s College London), Ruba Abu-Salma (King’s College London) and Jose Such (King’s College London) 

Bringing UX To The Table: Using Context To Turn The Art of UX Into a Science 

John N. A. Brown (School of Computing, Engineering and Technology, Robert Gordon University) 

19:30 – 22:30

Conference Dinner at South Hall Complex, Pollock Halls, University of Edinburgh 

Tuesday 8th July

Event Location: G.07, Informatics Forum, University of Edinburgh

9:00 – 9:40

Invited Talk by Bozidar Radunovic, Microsoft Research (Chair: Gareth Tyson)

The future of mobile cellular connectivity is programmable

In this talk, we will explore the potential of network programmability in cellular networks (5G and beyond), with a particular focus on radio access network (RAN). We will introduce a programmable radio and platform architecture, showcasing its capabilities, opportunities and challenges. Through examples, we will demonstrate how a programmable RAN can enhance network performance and enable new applications. We will also discuss how easy is to deploy this infrastructure and do your own research on top of it.

9:40 – 10:25

Session 4: Security & Privacy (Chair: Jingjie Li)

Ubiquitous Metadata Private Communication 

Alexandre Pauwels (University of Cambridge)

De-anonymising data in Wi-Fi positioning systems 

Michael Christian Fink Amores (University of Cambridge) and Alastair R. Beresford (University of Cambridge) 

Sideloading in iOS before, during and after the EU regulation 

Luis A. Saavedra (University of Cambridge), Alastair R. Beresford (University of Cambridge), Alice Hutchings (University of Cambridge) and Hridoy S. Dutta (University of Cambridge) 

10:25 – 10:55

Coffee Break

10:55 – 11:35

Invited Talk by Roeland Nusselder, Plumerai (Chair: Nic Lane)

Towards trillions of Tiny AIs on the edge

Already running on millions of devices in the field and rapidly growing, Plumerai’s Tiny AI is fast, cheap, battery-powered, and highly accurate. Through deep integration of all layers of the software stack—from the inference engine and datasets to model architecture and pre- & post-processing—it achieves unbeaten accuracy with a tiny memory footprint. With its growing presence on the edge, Plumerai is now connecting multimodal LLMs to embedded devices to unlock radically new applications.

11:35 – 12:50

Session 5: Edge AI (Chair: Mirco Musolesi)

On Harnessing Idle Compute at the Edge for Foundation Model Training 

Leyang Xue (University of Edinburgh), Eren Mendi (University of Edinburgh), Ismaeel Bashir (University of Edinburgh) and Mahesh K. Marina (University of Edinburgh) 

Junkyard Pre-Training: Federated LLM Pre-Training on Decentralized Edges 

Dongqi Cai (Beijing University of Posts and Telecommunications), Shangguang Wang (Beijing University of Posts and Telecommunications), Nicholas Lane (University of Cambridge) and Mengwei Xu (Beijing University of Posts and Telecommunications) 

FedTMOS: Efficient One-Shot Federated Learning with Tsetlin Machine 

Shannon How (University of Southampton), Jagmohan Chauhan (University College London), Geoff V Merrett (University of Southampton) and Jonathan Hare (University of Southampton) 

FedGuCci: Making Local Models More Connected in Landscape for Federated Learning at Edge 

Zexi Li (University of Cambridge), Dongqi Cai (University of Cambridge) and Nicholas Lane (University of Cambridge) 

EdgeAI Models for Human Activity Recognition on Low-Power Devices 

Vinamra Sharma University of Glasgow), Danilo Pau (STMicroelectronics) and José Cano (University of Glasgow) 

12:50 – 13:50

Lunch

13:50 – 14:30

Invited Talk by Suman Banerjee, University of Wisconsin-Madison (Chair: Steve Hodges)

Lightweight Edge AI for Sustainability, Public Safety, and More

 

Edge computing provides a new way to implement services with many unique advantages. While many edge computing solutions have been implemented within different network infrastructures, in this talk, we will explore ways to design a lightweight edge computing platform which is robust and portable, leading to interesting applications and services in sustainability, public safety, and many more application domains.

14:30 – 15:15

Session 6: Edge Computing (Chair: Yang Liu)

A Distributed Framework for Matrix Multiplication on a Cluster Raspberry Pis 

Rachel Somerset (University of Edinburgh) and Yuvraj Patel (University of Edinburgh) 

RAPID-ASR: Runtime ASR Parallelization and Isolation on Mobile Devices 

Chong Tang (University of Southampton), Hao Dai (University of Southampton) and Jagmohan Chauhan (University College London) 

Edge-Based Anomaly Detection in Electric Vehicle Charging Infrastructure with Continual Learning 

Loic Lorente Lemoine (Cardiff University), Amanjot Kaur (Cardiff University), Nhat Pham (Cardiff University) and Omer Rana (Cardiff University) 

15:15 – 15:45

Coffee Break

15:45 – 16:45

Session 7: Sensing (Chair: Yingying Zhao)

MicroData: Mobile and distributed sensing with the micro:bit 

Kier Palin (Lancaster University), Steve Hodges (Lancaster University), Joe Finney (Lancaster University) and Thomas Ball (University of Washington) 

Whatever happened to SenseCam? 

Andrea Baumann (Lancaster University), Steve Hodges (Lancaster University) and Nigel Davies (Lancaster University) 

HAQSS: High-precision Adaptive Quantization Based on Sensor's Specification 

Pei Mu (University of Edinburgh, Yufeng Xia (University of Edinburgh) and Antonio Barbalace (University of Edinburgh) 

Where Has All The Smartphone Sensing Gone? 

Veljko Pejović (University of Ljubljana)

16:45 – 17:00

Closing and MobiUK 2026 Announcement