School of Computer Science, Bangor University
Opening remarks introducing the Future AI 2026 conference, outlining key themes across AI research, industry collaboration, public services, and innovation in Wales and beyond.
Professor Jonathan C. Roberts is overall chair and organiser of Future AI, 2026.
Professor Jonathan C. Roberts is a leading academic in visualisation and human‑centred computing (HCC) at Bangor University, working within the School of Computer Science and Engineering. He is internationally recognised for his contributions to data visualisation, visual analytics, human–computer interaction, design‑thinking methodologies, computing pedagogy, and AI‑supported design. His research centres on the development of explanatory and exploratory human‑centred visual systems, focusing on how people analyse, visualise, and interact with data.
He is widely known as the creator of the Five Design‑Sheet methodology and as lead author of Five Design‑Sheets: Creative Design and Sketching for Computing and Visualisation (Springer Nature, 2017). Earlier in his career, he co‑developed ColloCaid.uk, a platform designed to help people improve their writing through the use of word collocations. His recent work integrates AI, particularly exploring how artificial intelligence can assist people in performing visualisation design studies more effectively.
NVIDIA
NVIDIA’s role in advancing AI research, infrastructure, and collaboration between academia and industry.
Andy Grant is the Director of Supercomputing, Higher Education, and AI for Europe, the Middle East, and Africa (EMEA) at Nvidia, bringing more than 25 years of high‑performance computing (HPC) expertise to the role. He has a distinguished track record in leading major HPC, AI, and quantum initiatives, having previously served as the Global Head of HPC, AI, and Quantum Sales at Atos. In that position, he oversaw large‑scale HPC programmes across multiple sectors, delivering numerous top‑50 HPC and major AI deployments. Earlier at Atos, he managed the HPC, big data, and cybersecurity division for the UK and Ireland, guiding the delivery of complex HPC and big‑data architectures.
Grant’s career also includes significant leadership roles at IBM, where he headed sales and business development for HPC in the UK and Ireland, managing both technical and commercial teams. He is a founding member of the UK Government’s e‑Infrastructure Leadership Council, contributing strategic guidance on national HPC and big‑data policy. Known for his enthusiasm for advancing the supercomputing and AI landscape, Grant continues to work closely with colleagues, partners, and customers across the EMEA region to drive innovation and support the next generation of large‑scale computing solutions.
Centre for Digital Public Services, Welsh Government
Overview of AI and digital transformation initiatives within Welsh public services.
Peter leads the learn digital skills service to build digital transformation skills and confidence of staff in existing roles and support those who want to move into Welsh digital public services.
He works with others to develop a digital workforce strategy and support practical measures to create a pipeline of skilled professionals with the goal of reducing digital skills gaps, build a shared understanding of digital transformation and user-centred design, upskill existing Welsh public sector employees new to digital, and attract talent into digital roles in Wales.Pontus Stenetorp is with University College London (UCL), Karin Sevegnani is with NVIDIA
This talk explores collaborative approaches to building large language models in the UK, highlighting shared infrastructure, open research, and the benefits of coordinated national AI development.
Pontus Stenetorp is a Professor of Natural Language Processing in the Department of Computer Science at University College London (UCL), where he works at the intersection of NLP and machine learning. He also serves as Deputy Director of the UCL Centre for Artificial Intelligence and is the UCL–Facebook AI Research PhD Programme Director, contributing to leadership and strategy across UCL’s AI research ecosystem.
His research focuses broadly on large language models, machine reading, deep learning, and computational linguistics, areas in which he has published extensively and supervised world‑leading research teams. He has progressed through multiple academic roles at UCL—from Lecturer to Full Professor—and previously conducted research at the University of Tokyo and the National Institute of Informatics. [Homepage]
Karin Sevegnani is a senior solutions architect at NVIDIA, where she leads the NVIDIA AI Technology Centre (NVAITC) in the UK. In this role, she drives collaborations with higher education institutions and research organizations to advance AI innovation and adoption. Before joining NVIDIA, Karin worked as a research engineer in Edinburgh, applying her expertise in AI development. She holds a PhD in Conversational AI from a joint program between the University of Edinburgh and Heriot-Watt University, completed under a combined scholarship and degree arrangement. Her specialization lies in natural language processing (NLP), particularly in conversational AI systems. [Technical blog page]
Bangor University
An overview of language technologies, including speech and text processing, with particular reference to multilingual and Welsh-language innovation.
Gruffudd (Gruff) Prys is a Senior Terminologist and Head of the Language Technologies Unit at Canolfan Bedwyr, Bangor University. His work centres on Welsh‑language technology, terminology development, and digital language infrastructure, contributing extensively to national efforts in Welsh language standardisation and technological innovation.
He has played a key role in major terminology and language‑technology projects, co‑authoring influential resources such as Cysill Ar-lein, a corpus and toolset for contemporary Welsh spelling and grammar, and contributing to research on Welsh POS tagging, terminology management, and language technology in Wales. His publications span book chapters, research reports, and conference papers, reflecting a long‑standing commitment to strengthening Welsh in digital contexts.
NVIDIA
Jonny Hancox a Senior Solutions Architect from the Health and Life Sciences team at Nvidia, will talk on Medical imaging and the use of frameworks like MONAI.
Jonny Hancox a Senior Solutions Architect from the Health and Life Sciences team at Nvidia, will talk on Medical imaging and the use of frameworks like MONAI. He is an experienced Software Architect with a demonstrated history of working in both research and commercial environments. Strong skills in the design and development of innovative healthcare solutions using a variety of popular languages and frameworks. Solid record of working with Deep Learning for medical image analysis with a number of NHS departments and Pharmaceutical companies. [Technical blog page]
NVIDIA
Karin is a senior solutions architect at NVIDIA, who leads the NVIDIA AI Technology Centre (NVAITC) in the UK – will talk on NeMo framework for LLM and Agentic AI.
Karin Sevegnani is a senior solutions architect at NVIDIA, where she leads the NVIDIA AI Technology Centre (NVAITC) in the UK. In this role, she drives collaborations with higher education institutions and research organizations to advance AI innovation and adoption. Before joining NVIDIA, Karin worked as a research engineer in Edinburgh, applying her expertise in AI development. She holds a PhD in Conversational AI from a joint program between the University of Edinburgh and Heriot-Watt University, completed under a combined scholarship and degree arrangement. Her specialization lies in natural language processing (NLP), particularly in conversational AI systems. [Technical blog page]
NVIDIA
Ira Shokar works on Earth-2 and AI-Aided Engineering at NVIDIA and will provide a talk on AI for physical simulation, surrogate modelling and design optimization with PhysicsNeMo
Ira Shokar works on CAE and Earth-2 at NVIDIA, focusing on AI for physical simulation, surrogate modeling and design optimization with PhysicsNeMo. Before joining NVIDIA, he completed a PhD at the University of Cambridge Department of Applied Mathematics and Theoretical Physics, where he developed probabilistic deep learning emulators for chaotic and turbulent fluid flows. [Technical blog page]
NVIDIA
Paul is a Senior Solutions Architect at NVIDIA and will talk on accelerated compute/HPC approaches.
Paul Graham joined NVIDIA in Dec 2018 where he has responsibility for supporting customers and partners in delivering accelerated solutions for the higher education, high performance computing, and AI communities in the UK. Previously, he spent 20 years working at EPCC, the supercomputing center at the University of Edinburgh, where he worked on a broad range of academic and industrial projects, parallelizing and optimizing code, and assisting small and medium businesses in gaining access to and using HPC facilities. As well as providing advice to researchers on making the best use of their NVIDIA hardware, Paul teaches how to program for GPUs, mentors at hackathons, and regularly engages with the research software engineering groups at UK universities. [LinkedIn> | Technical blog page].
NVIDIA
This session will introduce NVIDIA’s approach to accelerating applied AI through its technology framework, highlighting how GPU‑powered pipelines support modern workflows across domains such as large language models, data processing, and model deployment. Karin Sevegnani will outline key components of the NVIDIA ecosystem; spanning software, tools, and collaborative platforms; and discuss how these resources enable researchers and developers to build and scale AI systems more efficiently.
Karin Sevegnani is a senior solutions architect at NVIDIA, where she leads the NVIDIA AI Technology Centre (NVAITC) in the UK. In this role, she drives collaborations with higher education institutions and research organizations to advance AI innovation and adoption. Before joining NVIDIA, Karin worked as a research engineer in Edinburgh, applying her expertise in AI development. She holds a PhD in Conversational AI from a joint program between the University of Edinburgh and Heriot-Watt University, completed under a combined scholarship and degree arrangement. Her specialization lies in natural language processing (NLP), particularly in conversational AI systems. [Technical blog page]
Bangor University – Vice Chancellor
Welcome address highlighting Bangor University’s strategic commitment to AI research, innovation, and regional economic development.
Professor Edmund Burke is the Vice-Chancellor of Bangor University. With a career dedicated to higher education and the pursuit of academic excellence, Professor Burke has held senior positions at the University of Nottingham, University of Stirling, Queen Mary University of London as well as, most recently, at the University of Leicester. He has a distinguished international research profile in Operational Research. His research investigates intelligent decision support methodologies in complex environments and lies at the interface of Computer Science and Mathematics. He is a Fellow of the Royal Academy of Engineering and President of the Operational Research Society.
Bangor University
This presentation provides an overview of AIMLAC (the Artificial Intelligence, Machine Learning and Advanced Computing doctoral training centre) and highlights some of the School of Computer Science and Engineering's Human‑Centred Computing (HCC) and AI initiatives. The talk explores how Bangor is advancing inclusive design, interdisciplinary research, and AI‑focused skills development, ensuring that emerging technologies are shaped around people, cultures, and communities.
AIMLAC supports a new generation of researchers at the intersection of AI and advanced computing, having funded 12 fully supported PhD positions. These projects span a wide range of innovative computational challenges, including Large Language Models and Law, Animal Re‑identification in Video, Semi‑supervised Species Identification, and Grammatical and Ecological Neuroevolution. Together, they illustrate how Bangor researchers apply AI to meaningful real‑world problems.
The presentation will also introduce Get Into AI, an EPSRC‑funded Landscape AI Skills Programme led by Bangor University and launched in October 2025. The project aims to increase awareness, engagement, and diversity in AI skills, especially among schools in North Wales, while also reaching wider audiences through public events and online resources.
Get‑Into‑AI pilots innovative, hands‑on approaches to AI literacy and engagement, combining physical computing, creative coding, playful learning, and public outreach. Its eight work packages include AI in Welsh‑medium schools, AI starter kits based on Micro:bit, community creative‑coding workshops, an online hub of open educational resources, LEGO‑based AI demonstrations, and a student hackathon.
Together, AIMLAC and Get‑Into‑AI demonstrate Bangor University’s commitment to building a responsible, inclusive, and human‑centred AI ecosystem—one that develops cutting‑edge research while broadening opportunities for learners and communities across Wales.
Professor Jonathan C. Roberts is overall chair and organiser of Future AI, 2026.
Professor Jonathan C. Roberts is a leading academic in visualisation and human‑centred computing (HCC) at Bangor University, working within the School of Computer Science and Engineering. He is internationally recognised for his contributions to data visualisation, visual analytics, human–computer interaction, design‑thinking methodologies, computing pedagogy, and AI‑supported design. His research centres on the development of explanatory and exploratory human‑centred visual systems, focusing on how people analyse, visualise, and interact with data.
He is widely known as the creator of the Five Design‑Sheet methodology and as lead author of Five Design‑Sheets: Creative Design and Sketching for Computing and Visualisation (Springer Nature, 2017). Earlier in his career, he co‑developed ColloCaid.uk, a platform designed to help people improve their writing through the use of word collocations. His recent work integrates AI, particularly exploring how artificial intelligence can assist people in performing visualisation design studies more effectively.
Cardiff University, Technical Director of Supercomputing Wales
An overview of high-performance computing infrastructure in Wales and its role in accelerating AI research and industrial innovation.
We present an overview of the federated Supercomputing Wales (SCW) services, the national research facility for Wales delivered by a consortium of Aberystwyth, Bangor, Cardiff and Swansea Universities. Focusing initially on the high quality supercomputing-enabled scientific research delivered by SCW, consideration is given to the sustainability challenges now being addressed by the service.
In assessing the evolution of research computing, our focus lies in the increasing demand for hybrid double-precision (FP64) scientific computing and AI workload acceleration. Such systems combine the high numerical precision required for rigorous physics simulations with the tensor processing power needed for Deep Learning, essential in unlocking a new class of simulation-led AI research in, for example, Chemistry, Materials and Computer Science. Whether this is best addressed through native FP64 or emulated FP64 remains an open question.
Professor Martyn F. Guest is Director of Advanced Research Computing at Cardiff University. He obtained his B.Sc. in Chemistry from Sussex University in 1967 and his Ph.D. in Theoretical Chemistry, also from Sussex, in 1971 under the direction of Prof. J.N. Murrell. Following a postdoctoral position with Prof. I.H. Hillier at the University of Manchester, Professor Guest joined the Science and Engineering Research Council in 1972, first at the Atlas Computer Laboratory (Chilton, Oxfordshire), and from 1979 at the Daresbury Laboratory near Warrington.
He spent three years as Senior Chief Scientist and Technical Group Leader of the High Performance Computational Chemistry Group at the Pacific Northwest National Laboratory (Richland, Washington, USA). Professor Guest returned to the UK in 1995 as Associate Director of the Computational Science and Engineering Department at the STFC Daresbury Laboratory (UK) before taking up his present position at Cardiff University in Mach 2007.
At M-SParc, we’ve been building a very practical model of applied AI—one that starts with real-world problems faced by SMEs. One example is Egni.ai, our AI-powered carbon reporting tool. It combines DEFRA emissions factors with large language models to translate complex carbon accounting into clear, auditable carbon footprint reports that SMEs can actually understand and act on. What’s important here isn’t just the tech—it’s accessibility. We’re lowering the barrier to net zero by making compliance, reporting, and decision-making usable for everyday businesses, not just specialists.
That model is now evolving. Having proven its value with SMEs across Wales, we’re pivoting Egni.ai into new, data-rich sectors such as sport and agriculture—where emissions data, behavioural change, and real-time insight can unlock huge gains. In agriculture, this means connecting farm-level activity with credible carbon metrics. In sport, it’s about understanding the footprint of clubs, facilities, and events. In both cases, the opportunity is the same: pairing trusted datasets with AI reasoning to move from static reporting to continuous insight—and doing so in a way that is scalable, secure, and grounded in real operational contexts.
Alongside this, we’re leading national research into how Welsh businesses are actually adopting AI. Working with Welsh Government, we’re surveying SMEs across sectors to understand readiness, use cases, risks, skills gaps, and good practice. This is not abstract research—the findings will directly inform a pan-Wales programme of AI support for SMEs. The aim is to ensure adoption is responsible, inclusive, and economically meaningful, while avoiding the familiar trap of AI being either over-hyped or under-used.
Finally, M-SParc acts as a living lab for good AI in action. A great example is 42able.ai, whose work in the education sector shows how AI can be deployed ethically, transparently, and at scale to improve outcomes. These kinds of collaborations—between research, industry, and applied infrastructure—are where we see the real opportunity. Wales may be small, but that gives us an advantage: we can connect compute, research, policy, and SMEs into a single, testable ecosystem. That’s the model we’re building—and the one we believe can travel well beyond Wales.
Pryderi has been providing strategic direction as Director of M-SParc since 2018. His leadership is focused on the vision: to ensure that local communities, the economy, culture, and the environment can flourish sustainably through the creation of high-quality employment opportunities. With a background in local government and extensive entrepreneurial experience since the age of 16, Pryderi is dedicated to fostering an environment where businesses and communities can thrive. Outside of work, Pryderi enjoys watching Arsenal, cycling, and listening to music of all kinds. He is also a keen socialiser(!) and runs a pizza van and a glamping business!
Albert Gubay Business School, Bangor University
This talk explores how machine learning models can enhance investment decision-making and trading strategies through predictive analytics and data-driven optimisation.
Bruce’s work focuses on using data-driven predictive modelling techniques to solve difficult business problems. His work has been applied in fields such as health, logistics, finance and marketing. He is also particularly interested in finding better ways to apply machine learning techniques, statistics and data science to investment and trading.
Richard Scott .net
Examines organisational AI adoption, the risks of shadow AI, and the growing digital skills gap facing modern enterprises.
This talk explores how AI is actually being adopted across organisations today — not as a future roadmap item, but as a present-day reality. Delivered in partnership context with M-SParc and NVIDIA, the talk focuses on AI as a capability and skills shift, rather than a purely technical challenge. The presentation is narrative-led and sector-agnostic, designed for a mixed audience of founders, SMEs, technologists, public-sector leaders, and innovators.
The talk starts by framing AI as a step-change in how intelligence is accessed and applied. The rapid acceleration of AI capability explains both the opportunity and the unease many organisations are experiencing. AI is briefly demystified through a simple evolutionary lens, giving the audience shared language and confidence without technical depth. The talk reframes AI as a set of reusable capabilities that augment human thinking and decision-making, rather than a collection of standalone tools. With this foundation, the session introduces Shadow AI — the informal, often invisible use of AI by individuals and teams. This is positioned as a natural response to accessible intelligence and unmet skills needs, not a failure of governance. Most AI risk and missed value is shown to stem from gaps in literacy, confidence, and understanding — not from the absence of platforms or policies. The talk concludes with a human-centred approach to AI adoption: build understanding first, enable safe experimentation, develop skills across roles, and allow strategy and tooling to follow.
Bangor University, Albert Gubay Business School
While AI research often focuses on algorithms and applications, organisations increasingly rely on narrative reporting to explain AI strategy, risks, and value creation. This talk explores how AI is transforming corporate storytelling and accountability.
Khaled Hussainey was appointed as a Professor in Accounting at Bangor Business School in April 2024. Formerly, he held positions at Portsmouth University, Plymouth University, Stirling University, Manchester University, and Ain Shams University. He has more than twenty-five years of teaching experience across undergraduate and postgraduate levels. He has a growing research reputation in the field of accounting and finance. His research interests focus on corporate narrative reporting. He has been featured in the list of “World Ranking of Top 2% Researchers” database created by experts at Stanford University (USA). He has published more than 200 articles published in peer-reviewed journals and has written a number of book chapters and edited several books. His articles have appeared in such journals as Accounting and Business Research, British Accounting Review, Business Strategy and the Environment, European Financial Management, International Journal of Finance and Economics, International Review of Financial Analysis, Journal of Accounting and Public Policy, Journal of Accounting Literature, Journal of Corporate Finance, Journal of Environmental Management, Journal of International Accounting, Auditing and Taxation, Review of Quantitative Finance and Accounting, and Technological Forecasting & Social Change. He has received external research grants from Economic and Social Research Council (ESRC); British Academy; Qatar National Research Foundation (QNRF), Financial Reporting Council (FRC) and Abu Dhabi University. He is Co-Editor of Journal of Financial Reporting and Accounting and Editor of International Journal of Emerging Markets. He is Associate Editor of International Journal of Finance and Economics, Journal of Applied Accounting Research, International Journal of Accounting, Auditing and Performance Evaluation and Review of Accounting and Finance.
Bangor University, Albert Gubay Business School
This work presents how artificial intelligence is shaping modern business strategy and entrepreneurship. It explores how organisations and entrepreneurs can use AI not just as a tool, but as part of their wider strategic planning and innovation processes.
Dr Siwan Mitchelmore is a Senior Lecturer in Business and Management at Bangor Business School and is an expert researcher in Entrepreneurship, Leadership and Innovation in SMEs. Siwan's work has appeared in many International recognised Journals including Journal of International Management, Journal of Business Research and Journal of Small Business Management. Siwan was recognised for her excellence in teaching and was awarded a Bangor University Teaching Fellowship in the the 2016 University Graduation Ceremony.
Science and Technology Facilities Council (STFC), UK
This work presents gVirtualXray (gVXR), a GPU‑accelerated simulation framework X-ray imaging. It can produce a high number of realistic simulated images to train machine learning algorithms, including deep neural networks, which would not have been possible a few years ago. Early examples of applications include (i) the estimation of the length statistics of aggregate fried potato products, (ii) the creation of generative adversarial networks to generate X-ray radiographs from photographs of hands, and (iii) the study of real-time edge-aware denoising in fluoroscopic devices. Nowadays, gVXR is widely used in data augmentation, particularly for image segmentation tasks such as (i) lung nodule detection in 2D radiographs, (ii) segmentation of woven carbon fibre reinforced woven composite, (iii) material decomposition in spectral CT data, (iv) defect detection and characterisation and (v) characterisation of surface roughness for additive manufacturing.
Franck is a Computational Scientist in Imaging at the Science and Technology Facilities Council (STFC), the research arm of UK Research and Innovation (UKRI), and Honorary Professor at Bangor University. His research focuses on image analysis and computer vision, mathematical optimisation, and high-performance computing. He is particularly interested in X-ray Computed Tomography (CT), investigating how high-performance computational simulations can improve experimental X-ray imaging.
Franck has held research positions in France, Wales and the United States of America and developed a keen interest in large multi-disciplinary research projects. He completed his MRes in Imaging and Systems at Institut National des Sciences Appliquées de Lyon (INSA-Lyon), France, in 2003, working in close collaboration with the European Synchrotron Radiation Facility (ESRF), a pan-Europe large scale particle accelerator. He mathematically modelled the experimental response of X-ray detectors to remove artefacts in 3D tomography. Franck achieved his PhD research in the School of Computer Science at Bangor University in 2008. His research area addressed medical virtual environments, particularly for interventional radiology training. He worked on a simulator for ultrasound guided needle puncture: BIGNePSi (Bangor Image Guided Needle Puncture Simulator). Following up his PhD research, he worked as a research officer in the School of Computer Science at Bangor University.He was the technical lead of a large multi-disciplinary project on the implementation and validation of ImaGINe-S (Imaging Guided Interventional Needle Simulation), a VR Simulator for Visceral Needle Puncture. ImaGINe-S was awarded 2nd place in the Eurographics 2009 Medical Prize. In late 2008, he took a secondment position at INRIA Saclay—Île-de-France (the French national research institute in computer science and automation) and the French Atomic Commission where he was investigating novel tomography reconstruction algorithms in nuclear medicine. He then worked at the Department of Radiation Oncology of the University of California, San Diego (UCSD) where he developed Compton scattering computations on GPU. As part of his position, Franck had clinical duties including the monthly testing and calibration of the medical particle accelerators. In 2011, Franck went back, 1st to INRIA Saclay—Île-de-France to investigate multimodal visualisation of MRI data, 2nd to Bangor University to take up a lectureship position in Computer Science. He was promoted to Senior Lecturer in 2018 and Reader in 2023.School of Ocean Sciences, Bangor University, Menai Bridge, UK
Use of edge sensors and AI systems to monitor marine wildlife, transforming raw environmental data into actionable conservation insights.
Understanding marine ecosystem functioning requires detailed knowledge of the cryptic behaviours and interactions of marine predators. Bio-logging technologies (animal-borne sensors) are increasingly deployed to monitor animal movements, foraging behaviour and environmental context. However, these technologies generate large, complex and heterogeneous datasets (multi-dimensional signals sampled at up to 10,000 Hz) that demand robust analytical frameworks and advanced Artificial Intelligence (AI) tools. Here, I present two complementary AI- driven case studies in marine ecology: high-frequency scalar sensor data and underwater image data, highlighting opportunities for multimodal modelling and edge AI. First, high-resolution accelerometer tags generate multi-dimensional time-series data that provide detailed proxies for behaviour but are difficult to interpret at scale. I evaluated unsupervised and supervised machine-learning approaches on multi-season accelerometer datasets from penguins, explicitly accounting for inter-individual variability. Expectation Maximisation clustering was used to generate behavioural labels, which were then used to train Random Forest classifiers. Incorporating behavioural variability resulted in high agreement (>80%) between approaches and consistent energy expenditure estimates, although behaviours with similar signal signatures remained difficult to distinguish, underscoring challenges in scalable behavioural inference.
Second, emerging image-based bio-logging technologies, including animal-borne cameras and active acoustic systems, provide contextual data on predator–prey interactions and environmental conditions. Underwater imagery poses unique computer vision challenges such as variable illumination, colour distortion, motion blur, dynamic backgrounds and scale variability. I review state-of-the-art computer vision tools for underwater object detection and tracking, propose a four-phase processing framework (video manipulation, image processing, image labelling and model development), and discuss integration with scalar sensor data collected at different temporal resolutions. I further explore lightweight on-board AI models for real-time inference at the sensor level, enabling edge processing to reduce data storage, transmission costs and latency. I argue that integrating AI across scalar and image-based bio-logging, using multimodal models, GPU-accelerated pipelines and edge computing, will transform marine ecological research and conservation. Such frameworks enable scalable behavioural inference, real-time ecological monitoring and data-driven conservation decision-making, creating opportunities for collaboration between ecologists, AI researchers and industry partners in high-performance and embedded AI systems.Marianna is a quantitative ecologist driven by a passion for advancing the fields of ecology, movement ecology, habitat conservation, and species monitoring. She obtained her PhD at the University of Aberdeen in Scotland, where she investigated the foraging behaviour of two diving seabird species—common guillemots (Uria aalge) and razorbills (Alca torda)—during their breeding season. Using combined tracking data from GPS, Time Depth Recorders, and accelerometers, she characterised their foraging strategies both above and below the water, identifying common patterns, differences in movement, and distinct hunting strategies. Since completing her PhD, Marianna has worked with a diverse range of species, including Arctic and Northern Hemisphere seabirds, Arctic terrestrial mammals, Mediterranean marine turtles, and even Antarctic wildlife. Her research broadly focuses on questions related to animal movement processes, responses to environmental change, and the links between movement, population dynamics, and life‑cycle strategies. She uses biological data—such as accelerometers, GPS tags, and SRDL devices—to understand how, where, when, and why animals move. Marianna is equally enthusiastic about coding and modelling, frequently working with statistical tools and individual‑based models (IBMs). At the same time, she thrives in the field—nothing brings her more joy than hands‑on ecological work outdoors. She is joining the School of Ocean Sciences at Bangor University as Lecturer in Marine Top Predators Ecology, where she looks forward to continuing her research on how changes in the availability of marine resources affect the energetics, fitness, behaviour, decision‑making, and habitat use of marine predators.
School of Computer Science and Engineering, Bangor University
This work explores how visible light can not only be used for illumination and communications, but also as a highly accurate indoor positioning system. By combining signal characteristics from light sources with advanced machine learning techniques, the approach significantly improves both 2D and 3D location accuracy. The talk highlights methods for extracting meaningful features from light signals, training predictive models, and addressing challenges such as noise, multipath effects, and dynamic environments.
Aran White is an MEng Electronic Engineering student and a Junior Researcher at the DSP Centre. His work focuses on Visible Light Positioning, Optical Wireless Communication technologies, and next‑generation indoor navigation systems, where he contributes to developing advanced signal‑processing solutions for future smart‑environment applications.
Bangor University
AI is becoming an increasingly common part of modern game development, including how we animate characters. Procedural animation spans classic algorithmic techniques through to newer AI-driven approaches, generating dynamic, real-time movement that can bring characters to life. In this talk I'll explore how I use procedural animation in my VR and mixed-reality work to drive immersive creature motion and behaviour — from playful dolphins to fearsome aliens. I will also explore what generative AI means for game development today, highlight key industry trends, and consider how the process of making games could evolve in the future.
Dr Llŷr ap Cenydd is a VR developer and Lecturer in the School of Computer Science and Engineering at Bangor University. His work focuses on virtual reality, real‑time graphics, procedural animation, and artificial life. Using AI and smart decision, he develops interactive and immersive games. He is best known as the creator of Ocean Rift, one of the earliest and most influential underwater VR experiences, allowing users to explore marine environments and encounter dolphins, whales, sharks, and other ocean life. Ocean Rift website: https://oceanriftvr.com/.
Llŷr is also the solo developer behind Crashland, an action‑horror VR game praised for its fast combat, advanced procedural animation, and immersive creature design. Crashland website: https://www.crashlandvr.com/
Alongside his commercial VR work, Llŷr teaches game design and development and continues to research cutting‑edge VR technologies, including procedural animation and interactive simulation.
School of Arts, Culture and Language, Bangor University
This talk considers how AI might create meaningful learning environments where complexity, choice, and creativity can flourish.
Eben researches the cultural, social, and spatial significance of bookstores—both physical and digital—as meaningful locations where ideas circulate, communities gather, and new possibilities emerge. His work spans interviews, corpus analysis, and cultural studies, including The Fantasy of the Bookstore (2022), and he supervises postgraduate research on indie bookselling, book culture, and the wider book trade.
Education, Bangor University
This talk will explore how AI can be made more accessible for learners with dyslexia, drawing on insights from the Get Into AI project and recent research in inclusive education. It will highlight practical approaches for designing AI‑supported learning that supports neurodivergent students and widens participation.
Sophie Todd graduated from Bangor University with a first-class honor’s degree in Psychology in August 2021. During my degree, I was able to gain experience tutoring children with dyslexia one to one through the ‘leap into literacy program’. She went on to graduate from Durham University with her PGCE in primary education, and recently completed her MA in Education at Bangor University, where she focused on dyslexia, accessibility, and inclusive learning. She has been working as a primary school teacher in Cyngor Sir Ynys Mon, Isle of Anglesey County Council. Recently she has been an intern on the Get Into AI project, specialising in making AI‑related educational resources more accessible to learners with dyslexia.
Her work sits at the intersection of education, accessibility, AI, and literacy, with a strong commitment to improving learning experiences for neurodivergent students.
School of Arts, Culture and Language, Bangor University
This talk will explore how AI can can be used in Mediaval Literature, and within postgraduate research.
Professor Niebrzydowski is a professor of Medieval Literature. She has been Director of the Graduate School for the College of Arts and Humanities during her career at Bangor University. Sue Niebrzydowski is a director of The Stephen Colclough Centre for the History and Culture of the Book at Bangor.
School of Computer Science and Engineering, Bangor University
Artificial intelligence has the potential to transform how we learn, work, and solve complex problems. As a Computer Science student, I will reflect on how AI enhances learning and innovation, while also raising important challenges around understanding, ethics, and responsible use.
Steph Parry is a third year student studying for her BSc in Computer Science.
School of History, Law and Social Sciences, Bangor University
This presentation explores the implications of Artificial Intelligence for today’s law students as future legal professionals. As AI becomes embedded in legal research, drafting, dispute resolution, and advisory work, it is reshaping the skills and responsibilities lawyers will need. The presentation considers how AI may transform access to justice, both expanding affordable legal services and risking digital exclusion and algorithmic bias, while also highlighting the social and environmental impacts of data-intensive technologies. It argues that law students must develop technological literacy alongside strong ethical judgment, critical thinking, and social awareness. Preparing for an AI-enabled profession means not only learning to use new tools, but understanding their broader consequences for fairness, sustainability, and the rule of law.
Simon Goy is a third-year law student, enrolled in the LLB, English law and French law at Bangor University. He is passionate about international and European public law, and is looking to pursue a career in the European institutions. He has a particular interest in issues relating to the impact of new technologies on users' rights, and the way of regulating these new technologies.
The work is presented under the supervision of Dr Sarah Nason, Law.
School of History, Law and Social Sciences, Bangor University
In this work, we will set out how different AI technologies have found their way into the criminal justice system, from facial recognition and predictive policing to automated case summaries in the courts, and automated risk assessments tools in probation. It will reflect on the differing governance schemes proposed across Europe and England and Wales and highlight some of the benefits promised as well as concerns raised about the largely unsystematic adoption of AI technologies in the criminal justice arena.
Martina studied law at the University of Tübingen before completing an MSc in Criminology and Criminal Justice at the University of Edinburgh in 1999, and a DPhil on the influence of the media on public perceptions of crime and criminal justice at the University of Oxford in 2008. She worked as a research officer at the Centre for Criminology, University of Oxford for six years on numerous projects funded by the Youth Justice Board, the Home Office, and the Nuffield Foundation.
Martina joined the Bangor University in 2007 and has since undertaken a range of research projects relating to the workings of the criminal justice system, including policing, the changes to probation, and questions of penal policy. Martina is a member and Co-Director of the ESRC funded WISERD Civil Society Centre; and is a Co-Director at the Welsh Centre for Crime and Social Justice.Bethan is currently Reader in Criminology and Criminal Justice. She joined the School of Social Sciences in January 2016, having previously been the recipient of several grants and fellowships at the School of Law, University of Manchester (2011-2015), and the Centre for Criminology and Criminal Justice, University of Oxford (2008-2011). Bethan holds a PhD in Criminology (Keele University), an MA in Comparative Criminology and Criminal Justice (Bangor University, received with distinction) and a BA in Criminology and Criminal Justice (Bangor University, first class).
Her research interests lie in the areas of policing and security, in particular: policing cultures, covert/undercover policing and surveillance; the operation and governance of private security; border enforcement; the use of Artificial Intelligence in law enforcement, and the policing of social divisions. She has conducted two major ethnogra phic field studies, both of which involved conducting prolonged observations of police officers as they went about their ordinary duties. She has been awarded research grants from the ESRC and Leverhulme Trust, and won a Simon Fellowship at the University of Manchester where she conducted research on the policing of international borders. Bethan is the author of Police Culture in a Changing World (Oxford University Press) and articles in major journals.
School of Computer Science and Engineering, Bangor University
This talk will focus on developing appropriate human-centered computing applications for law. The work will focus on several projects. First how visualisation can help to display and explain processes within public law and access to justice. Second how LLMs can help to recommend and report on suitable resources from specific law-based questions, using a RAG technique. The talk will discussion of regulatory, ethical, and governance challenges arising from AI deployment and human centered computing.
Dr Peter W. S. Butcher is a Lecturer in Human‑Computer Interaction (HCI) in the School of Computer Science and Engineering at Bangor University. His research focuses on immersive analytics, data visualisation, and web‑based XR (VR/AR), developing interactive systems that help users explore and understand complex data in immersive and situated environments.
He holds a PhD in Computer Science from the University of Chester, following an MSc and BSc at Bangor University, and has contributed to multiple high‑impact research projects across immersive analytics, collaborative visualisation, and cross‑disciplinary HCI applications. His work has appeared in prestigious venues such as IEEE Transactions on Visualization and Computer Graphics.
Peter teaches across advanced computing topics, including High‑Performance Computing (HPC) and advanced computational methods, drawing on his background in web‑based XR systems and data‑centric engineering. He is also engaged in interdisciplinary research on AI in Law, collaborating with Dr Sarah Nason (School of Law, Bangor University) to explore how artificial intelligence, visualisation, and decision‑support systems can support legal processes, public‑law analysis, and access to justice.
He is part of Bangor’s Immersive Environments Laboratory and contributes to research within the DSP Centre, while supervising student projects in immersive analytics, HCI, VR/AR, data visualisation, HPC and AI.
Glory Ogbonda is a data scientist and a full-stack developer with a passion for creating data-driven solutions that drive business results and social impact. He has years of experience in analysing complex data sets, developing predictive models, and designing interactive visualisations using a wide range of technologies and techniques, such as machine learning, data visualisation, and statistical analysis. Currently, is pursuing a PhD in Computer Science at Bangor University, as part of the UKRI CDT in Artificial Intelligence, Machine Learning and Advanced Computing (AIMLAC) programme where he is conducting a research on the use of AI in visualising the UK public law.
Bangor University
This talk explores cognitive and technical perspectives on AI hallucinations, examining reliability, interpretation, and human trust in machine-generated outputs.
Professor Oliver Turnbull is the Deputy Vice-Chancellor and Professor of Neuropsychology at Bangor University. He is an internationally recognised neuropsychologist whose research focuses on the role of emotion in mental life, with particular interests in emotion‑based learning and intuition, emotion in delusional beliefs, the neuroscience of psychotherapy, emotion regulation, and emotional memory, including its preservation in profound amnesia.
Alongside his academic work, Professor Turnbull is an experienced clinician, working with patients who have neurological lesions—especially those recovering from stroke and traumatic brain injury. He is the author of around 200 scientific publications, and co‑author of the widely translated popular science book The Brain and the Inner World (with Mark Solms), as well as Mistakes in Clinical Neuropsychology (with Christian Salas and Rudi Coetzer).
School of Computer Science and Engineering, Bangor University
Exploration of strategic regional growth initiatives, AI infrastructure development, and long-term innovation planning.
Michael Rushton is a Professor in Nuclear Energy at Bangor University, working within the Nuclear Futures Institute. His research focuses on atomic‑scale simulation of energy materials, with particular expertise in nuclear fuel, wasteforms, and actinide oxides. He specialises in developing high‑fidelity potential models, and in using molecular dynamics and lattice‑based methods to understand fuel behaviour—especially the interactions between fission gas bubbles and nuclear fuel lattices.
Before joining Bangor in 2017, Professor Rushton worked at Imperial College London in the Centre for Nuclear Engineering, where he helped establish Bangor’s BWR Hub and Network, on which he continues to serve. His work spans both crystalline and amorphous materials, with applications in nuclear fuels, batteries, and fuel‑cell materials. He has also contributed to major international programmes, including the EPSRC UK–Japan collaboration on radiolytic heating of high‑dose adsorbents and the NFIR fuels programme for EPRI.
Alongside his research, Michael regularly engages in public communication on nuclear technologies, including panel contributions at COP Cymru on nuclear co‑generation and media commentary on energy security
School of Computer Science and Engineering, Bangor University
Overview of current research directions in intelligent agents and autonomous systems.
This talk will present Frontiers in AI, highlighting the ground-breaking research carried out by the AI and Intelligent Agents (AI:IA) Research Group at Bangor University, led by Dr. Bill Teahan. The AI:IA group has a long-standing track record of innovative contributions to Artificial Intelligence, particularly in the areas of language modelling, natural language processing, and intelligent systems. The talk begins with a brief history of the group, outlining its formation, evolution, and key achievements, including influential past work on compression-based language models and probabilistic approaches to text modelling that have helped shape modern NLP research.
The main focus of the talk is the group’s current and ongoing research, which spans a wide range of contemporary AI challenges. Key topics include advances in Large Language Models (LLMs), alternative compression-based language modelling paradigms, and the development of agentic AI systems capable of autonomous reasoning and decision-making. The talk will also explore research in natural language processing, machine translation, speech recognition, text summarisation and literature searching, demonstrating how these techniques are being applied to real-world problems.In addition, the presentation will highlight interdisciplinary work in biomedical text analysis and diagnosis, showcasing how AI-driven models can support data-driven decision-making in healthcare contexts. Throughout the talk, emphasis is placed on both theoretical foundations and practical applications, illustrating how the AI:IA Research Group continues to push the boundaries of intelligent systems research. The session concludes with a discussion of future research directions and emerging opportunities at the frontiers of AI.
Dr William J. Teahan is a Lecturer in the School of Computer Science at Bangor University. His research focuses on Artificial Intelligence, Intelligent Agents, and the application of text‑compression‑based language models to Information Retrieval (IR) and text mining, including Information Extraction.
Before joining Bangor, Dr Teahan held several international research posts. He was a Research Fellow in the Information Retrieval Group at The Robert Gordon University in Aberdeen (1999–2000), an invited researcher in the Information Theory Department at Lund University in Sweden (1999), and a Research Assistant in the Machine Learning and Digital Libraries Laboratories at the University of Waikato in New Zealand (1998). He completed his PhD at the University of Waikato in 1998, where his doctoral work focused on applying text compression models to the problem of modelling English text.
School of Computer Science, Bangor University
In these closing remarks, we reflect on the key themes emerging from the conference — innovation, responsibility, interdisciplinary collaboration, and practical implementation. Moving beyond experimentation with AI tools, the focus shifts toward thoughtful integration across education, research and industry. The session highlights the importance of leadership, capability-building and ethical foresight as AI systems become embedded in everyday practice. It concludes by outlining future opportunities for collaboration and sustainable AI development within and beyond the university.
Professor Jonathan C. Roberts is overall chair and organiser of Future AI, 2026.
Professor Jonathan C. Roberts is a leading academic in visualisation and human‑centred computing (HCC) at Bangor University, working within the School of Computer Science and Engineering. He is internationally recognised for his contributions to data visualisation, visual analytics, human–computer interaction, design‑thinking methodologies, computing pedagogy, and AI‑supported design. His research centres on the development of explanatory and exploratory human‑centred visual systems, focusing on how people analyse, visualise, and interact with data.
He is widely known as the creator of the Five Design‑Sheet methodology and as lead author of Five Design‑Sheets: Creative Design and Sketching for Computing and Visualisation (Springer Nature, 2017). Earlier in his career, he co‑developed ColloCaid.uk, a platform designed to help people improve their writing through the use of word collocations. His recent work integrates AI, particularly exploring how artificial intelligence can assist people in performing visualisation design studies more effectively.
Head of School, School of Computer Science and Engineering, Bangor University
Closing reflections on the Future AI 2026 conference and next steps for collaboration and growth.
Professor William Heath is the Head of School of Computer Science and Engineering, Bangor University. He is a specialist in feedback control theory, with research spanning both classical and modern digital control. His work focuses on bridging the clarity and intuition of classical control approaches with the computational power of digital implementations to address nonlinear control problems. A central theme of his research is the behaviour of systems with actuator nonlinearities, including saturation, rate constraints, backlash, and hysteresis.
He has a strong interest in multiplier theory within the framework of absolute stability theory, and his recent work includes developing new frequency‑domain characterisations of both continuous‑time O’Shea–Zames–Falb multipliers and their discrete‑time analogues. These theoretical advances support practical applications in model predictive control (MPC) and anti‑windup, where Professor Heath’s contributions help improve stability, robustness, and real‑world performance.