2025 International Seminar Series on Emerging Technologies in Cyber-Physical Systems (CPS)

Presentation

Presentation

Presentation

Presentation

Presentation

Presentation

Presentation

Presentation

Presentation

Presentation

Presentation

Presentation

Biography: Hi, I’m Vishal. I have over six years of university-level teaching experience in the UK and more than 16 years of overall teaching expertise. I hold a PhD in Cybersecurity, an MSc in Software Engineering, and a BE in Information Science. Currently, I serve as a Lecturer at Nottingham Trent University, following the Teaching and Research (T&R) pathway.

My core research interests include Cryptography, Information Security, Cybersecurity, and IoT deployment, with additional expertise in Computer Networks, the Internet of Things (IoT), SQL, MS Power BI, Algorithm Design, Data Structures, and AI/ML. I also enjoy coding in C, C++, and Python.

I have published numerous papers in prestigious journals, including IEEE Access, Elsevier, and Springer.

Title: Lightweight Cryptography for Resource Constrained IoT Devices

Abstract: The increasing adoption of IoT in safety-critical areas, such as smart factories and connected vehicles, presents significant security risks, making system dependability crucial. Traditional cryptographic methods are effective for resource-rich devices but unsuitable for constrained IoT environments, necessitating a lightweight alternative known as Lightweight Cryptography (LWC). Achieving an optimal trade-off between cost, performance, and security remains a challenge. The proposed AUM algorithm addresses this by incorporating a novel 5-bit S-box for strong security, a transpose-based permutation for efficiency, and an optimized key generation technique. Comparative analysis on the ASIC platform and cryptanalysis demonstrate that AUM outperforms existing models.

Biography: Professor Matthew Montebello is a distinguished academic and the Head of the Department of Artificial Intelligence at the University of Malta, where he plays a pivotal role in shaping the future of AI education and research. With a career that seamlessly blends computer science and education, he has developed a unique interdisciplinary expertise, specializing in the application of artificial intelligence to e-learning. His international influence is evident through his visiting academic status and adjunct professorship at the University of Illinois, Urbana-Champaign, where he collaborated on cutting-edge projects with both the Computer Science department and the College of Education. A pioneer in education technology, Professor Montebello has developed a next-generation Virtual Learning Environment (VLE) that integrates social media concepts, revolutionizing how education is delivered in the digital age. His commitment to professional development is reflected in his active involvement with the University of Malta’s Web Editorial Board, Digital Education Committee, and other key committees. Since November 2022, he has been at the forefront of addressing the impact of Generative AI and Large-Language Models on higher education, co-authoring guidelines and facilitating workshops to prepare academics and students for the evolving educational landscape. With a robust portfolio of publications and research in AI, e-learning, and computer science, he is widely recognized as a thought leader in these fields. Additionally, his extensive teaching experience, dating back to his early career in secondary schools, has allowed him to mentor countless students and junior academics, shaping the next generation of professionals in ICT and AI.

Title: The Role of Generative AI in Academia: Transforming Teaching, Assessment, Research, and Writing

Abstract: The talk will equip participants with the knowledge and skills to leverage Generative AI in their academic research and writing. Participants will explore AI-powered tools for literature reviews, references management, academic writing, and research dissemination while addressing ethical considerations, academic integrity, and limitations. The session includes practical advice, links to hands-on activities, and recommendation on integrating AI responsibly into their academic journey.

Biography: Prof. Dr. Imre Felde (Engineer of Information Technology) completed his PhD in 2007. He has been working for Bay Zoltán Foundation for Applied Research (BZAKA, Budapest, Hungary) as research fellow between 1995 and 2009. He was General Director of BZAKA between 2010 and 2012. He has been working as Vice Dean (2012-2018) at Obuda Univesity and as Vice rector (2019-2024). He was appointed Full Professor at the John von Neumann Faculty of informatics in 2022. He published more than 240 scientific and technical papers in field of engineering, computer science, mathemathics and process modeling.

Title: Analysis of mobility customs in urban area by processing mobile network data

Abstract: Mobile phones and even more, smart-phones, are now fundamental parts of our life, they are practically, always with us, wherever we go, almost as if they were a part of our body. The continuous communication between a device and the Mobile Phone Networks leaves traces at the Operator’s system of our mobility habits. Via these devices, the Mobile Phone Network can ‘sense’ our movements, which is the basis of the “Smart City” concept. In the last two decades show the potential of the human mobility characterization using Call Detail Records (CDR) in epidemiology, sociology and urban planning. The analysis of the human movement patterns on the basis of the CDR data, that makes it possible to examine a very large population cost-effectively resulted a number of discoveries about human dynamics.  In addition to, by mixing mobility indicators with external data, the hidden relationship between the daily customs and Socio-Economic Status (SES) of the dwellers could be evaulated.

Biography: Tom Gedeon is the Human-Centric Advancements Chair in AI at Curtin University and was recently the Optus Chair in AI. Prior to this, he was Professor of Computer Science and former Deputy Dean of the College of Engineering and Computer Science at the Australian National University. He gained his BSc (Hons) and PhD from the University of Western Australia.
Tom’s main research areas are Responsive and Responsible AI, and generative AI. His focus is on the development of automated systems for information extraction, from eye gaze and physiological data, as well as textual and other data, and for the synthesis of the extracted information into humanly useful information resources, primarily using neural/deep networks and fuzzy logic methods, and delivered in real, augmented and virtual environments.
Tom has over 400 publications, and has run multiple international conferences. He is a former president of the Asia-Pacific Neural Network Assembly, and former President of the Computing Research and Education Association of Australasia. He has been General Chair for the International Conference on Neural Information Processing (ICONIP) three times. He has been nominated for VC’s awards for postgraduate supervision at three Universities. He was a member of the Australian Research Council’s College of Experts 2018-2021, and continues from 2024-2026. He is an associate editor of the IEEE Transactions on Fuzzy Systems, and a member of the Governing Boards of the IEEE Systems Man and Cybernetics and the Asia-Pacific Neural Network Society.

Title: Responsive AI and Responsible AI for useful and privacy protected AI systems

Abstract: With the availability of low-cost sensors in today’s environment, we are increasingly cnapturing data directly from individuals’ behaviours via wearables and cameras. This enables us to create AI tools capable of reading human actions and reactions to the outcomes created by our AI systems, allowing us to enhance or adjust their outputs, and be responsive to the human. This closely resembles the nonverbal cues used by people during a conversation. Responsive AI refers to cutting-edge AI that responds to human actions and reactions to predict subtle emotional states. In practice today, these sensors include wearable devices that detect skin conductance, heart rate, muscular activation, and skin temperature, among other signals. Cameras are used to track eye gaze behaviours, capture videos, record thermal images, or utilise hyperspectral technology, and with a long term additional goal of replacing wearable sensors. The deployment of AI to discern nuanced human internal states introduces new privacy concerns in addition to the expected privacy implications associated with video cameras. These risks can be alleviated through the adoption of privacy-preserving techniques referred to as Responsible AI. This is a privacy-by-design method for managing the use of private and personal data. In practice, this means using adversarial generative algorithms to remove personally identifying data from sensor data streams and videos. We will discuss previous work in these areas to show how fully utilising Responsive AI needs the incorporation of Responsible AI principles.

Biography: Dr. Tamás István Józsa is a lecturer (equivalent to assistant professor) at the Centre for Computational Engineering Sciences at Cranfield University specialised in modelling and simulation (M&S). Cranfield University is an associate partner of GEMINI (https://dth-gemini.eu) with Tamás as institutional lead. This EU flagship project targets the advancement of stroke diagnosis and treatment based on M&S. Beyond biomechanics, Tamás collaborates with industrial and academic partners to tackle problems in internal and external aerodynamics, such as the optimisation of supersonic combustors and hypersonic waveriders, and the safety assessment of cryogenic hydrogen tanks and jet blasts. His educational activities spread across the Computational Fluid Dynamics (CFD), Aerospace Computational Engineering, and Computational Software Techniques in Engineering MSc courses.

Between 2022 and 2023, Tamás was a Career Bridging Fellow at Amsterdam University Medical Centres. Previosuly, he worked as a doctoral adviser and postdoctoral researcher at the University of Oxford. He completed his PhD at the University of Edinburgh in 2018 and earned an MSc in CFD at Cranfield University in 2014. He studied mechanical engineering at the Budapest University of Technology and Economics between 2008 and 2013 (BSc & MSc).

Title: Computational modelling of perfusion in health and ischaemic stroke

Abstract: Stroke claims approximately 6.5 million lives each year and more than 50 million people live with stroke-related disabilities worldwide. About 25% of stroke patients are less 65 years old. Average clinical trials associated with stroke treatment development are estimated to take three years and cost £25 million with success rate about 15%. Such trials pioneered chemical (thrombolysis) and mechanical (thrombectomy) procedures which are combined to remove the blood clots causing ischaemic strokes. However, treatment optimisation remains an outstanding challenge. For example, it is unclear whether certain patient groups could benefit from blood pressure medication, or from single instead of combined treatment.

Modelling and simulation methods are the workhorse of product engineering in every sector except medical device and drug development. Simulations emerge as a promising approach to reduce the resources needed for clinical trials, to optimise, and to personalise stroke treatments but several issues need to be resolved first. The presentation will provide the necessary anatomical and pathophysiological details to appreciate the difficulties of stroke modelling and treatment simulations. Thereafter, the key components of a comprehensive stroke simulation pipeline will be discussed. Focus will be placed on a porous continuum model of the entire human brain. Details of clinical data integration will be discussed which enables the simulation of hundreds of stroke patients. Verification, validation, uncertainty and error quantification activities will be shown.

Biography: Prof. Carl James Debono,obtained his B.Eng.(Hons.) degree in Electrical Engineering from the University of Malta, Malta in 1997 and Ph.D. degree in Electronics and Computer Engineering from the University of Pavia, Italy in 2001. In 1997 he was employed as a Research Engineer with the Department of Microelectronics at the University of Malta. In 2001 he was appointed Lecturer in Communications and Computer Engineering at the University of Malta and was promoted to Senior Lecturer in 2006, to Associate Professor in 2011 and to Professor in 2017. Prof. Debono has served as Deputy Dean of the Faculty of ICT between November 2009 and September 2015. He also served as Head of the Department of Communications and Computer Engineering between October 2015 and September 2019. He is currently the Dean of the Faculty of ICT. Prof. Debono has participated in a number of local and European research projects in the area of communication systems and image/video processing. His research interests are in Resilient Multimedia Transmission, Multi-view Video Coding, and Computer Vision.

Title: Use of Depth Information in Visual Communications and Processing Tasks

Abstract: Depth information can play an important role in enhancing visual communications and processing tasks across various applications. In this talk we will explore the use of depth information in 3D video coding, saliency detection, monitoring of persons and object tracking. By integrating depth data, we can achieve more accurate and efficient processing of visual-based tasks. 

Biography: Peter Galambos (Senior Member, IEEE) received his M.Sc. and Ph.D. degrees in mechanical engineering from the Budapest University of Technology and Economics (BME) in 2006 and 2013, respectively. From 2007 to 2008, he was a Research Intern at the Toshiba Corporate Research and Development Center. He then joined the Institute for Computer Science and Control of the Hungarian Academy of Sciences (MTA SZTAKI), where he held a Young Researcher Scholarship from 2010 to 2012. Between 2011 and the end of 2015, he served as a team leader at MTA SZTAKI, coordinating the development of the VirCA VR system and its research applications. In 2013, he joined Obuda University, where he has been actively involved in robotics-related research, development, and education. He is currently a Full Professor and Director of the Antal Bejczy Center for Intelligent Robotics at Obuda University, Budapest. Since July 2024, he has been serving as Vice-Rector for Innovation. His current research interests include Advanced Industrial Robotics and Control Systems, Cyber-Physical Systems, and Virtual Reality. He is the author or co-author of over 160 scientific publications, with more than 2,000 citations. Beyond his academic career, he is a successful entrepreneur, co-founder, and CTO of the MAXWHERE 3D Digital Twin software platform.

Title: Automated Synthetic Training Data Generation for Robot Vision

Abstract: In recent years, Deep Learning (DL) methods for visual tasks have achieved remarkable progress in both accuracy and robustness. Modern approaches increasingly enhance performance by leveraging multiple modalities or training DL models simultaneously across various tasks. Datasets with rich annotations and diverse modalities are essential, as they enable the development, training, and evaluation of more sophisticated and versatile models. This presentation introduces an innovative annotation methodology for generating synthetic datasets, extending the capabilities of the Blender Annotation Tool (BAT). BAT automates the generation of ground-truth data, including segmentation masks, depth maps, surface normals, and optical flow, within synthetic 3D scenes created in Blender—requiring minimal manual intervention. The lecture will provide an overview of the underlying context, followed by detailed application examples that demonstrate the effectiveness and flexibility of this approach.

Biography: John Abela is a resident professor in the Faculty of ICT at the University of Malta. He specializes in deep learning and natural language processing. His work explores the emergent properties of large language models and their scalability towards achieving AGI. With interests spanning AI, quantum computing, mathematics, and search and optimization, Professor Abela is dedicated to advancing the frontiers of computer science through innovative research and practical applications.  

Title: Generative AI and Large Language Models: Understanding Self-Attention

Abstract: This lecture offers an exploration of the evolution and inner workings of generative AI, focusing particularly on the development and impact of large language models (LLMs). We begin with a brief historical overview of deep learning, tracing its roots from early neural network research to the groundbreaking advances that set the stage for contemporary AI. Special attention will be given to the watershed moment marked by the 2017 Google paper that introduced Transformer models, revolutionizing natural language processing by supplanting previous architectures with self-attention mechanisms.

We then examine the rapid progression of LLMs through the lens of the ChatGPT series, beginning with ChatGPT 3.5 released in November 2022 and its subsequent iterations. This evolution is contextualized within the broader argument that emergent behavior arises as these models scale, underscoring why each generation of LLMs has grown in complexity and size. The lecture will elucidate the fundamental principles of self-attention, a mechanism that has not only enhanced performance but also enabled models to capture intricate contextual dependencies in language.

Further, the session integrates philosophical and technical perspectives by exploring Daniel Dennett’s views on the Turing computability of human intelligence, which challenge us to reconcile computational theories with the subtleties of cognition. In contrast, we will also present Geoffrey Hinton’s cautious stance on the rapid pace of AI advancement, reflecting on the ethical and safety implications that accompany these technological leaps.

This lecture is designed to provide undergraduate students with an understanding of the transformative developments in AI, while critically engaging with the theoretical and practical implications of emergent behaviors in large-scale models. Attendees will gain insights into both the technical intricacies and broader societal impacts of modern AI systems, preparing them to navigate and contribute to this rapidly evolving field.

Dr. Ali Safaa Sadiq
Associate Professor in Cyber Security and Director of the Cyber Security Research Group (CSRG)
Nottingham Trent University
ali.sadiq@ntu.ac.uk 

Biography: Ali Safaa Sadiq (Ali) is an Associate Professor in Cybersecurity and research leader of the Cyber Security Research Group in the Department of Computer Science at Nottingham Trent University. Ali is also a senior IEEE member and adjunct staff at Monash University and an honorary Associate Professor at the Centre for Artificial Intelligence Research and Optimisation, Torrens University Australia. Ali has served as a senior lecturer in Intelligent Networks at the University of Wolverhampton, and a lecturer at the School of Information Technology, Monash University, Malaysia. Previously he has also served as a senior lecturer at the Department of Computer Systems & Networking Department, Faculty of Computer Systems & Software Engineering, University Malaysia Pahang, Malaysia. Ali completed his first degree in Computer Science in 2004, after that Ali had 5 years of industrial experience in Computer Science and Networking. Ali had his MSc and PhD degrees in Computer Science in 2011 and 2014, respectively. Ali has been awarded the Pro-Chancellor Academic Award as the best student in his batch for both master’s and Ph.D. He has published several scientific/research papers in well-known international journals and conferences. He was involved in performing more than 10 research grant projects, where recently with the CSRG team could secured ~£5m research income to perform research projects related to cyber security innovation and development, including the UK national Power Grid supplier OFGEM. Ali has been also involved as a co-investigator with a research project CYBERMIND that was funded £91k by Innovate UK Cyber Academic Start-up Accelerator 2020. Also, he has led (PI) a funded research project called TrustMe, which is funded in two phases by Innovate UK and DCMS. The project creates an innovative new platform to help AI developers and data scientists add security, trust, and explainability to their AI-based decisions. The first phase has been funded with £31,338k, while the second phase was funded with over £60k to develop the proof of concept. Ali managed to develop a commercialised platform called TYMLO and launched a company named TYMLO Technology Ltd. Ali was also part of the project named SOLVD which was funded by the European Regional Development Fund (ERDF) that was supporting local SME companies related to cyber security. He has supervised more than 10 Ph.D. students and 30 Masters students as well as some other undergraduate final year projects. He is currently working on funded projects named Drive with Confidence: A Safe and Secure Driving System to Mitigate Remote Vehicle Hijacking Risks, and PRAVE: PRoactive Authentication and Verification Embedded Model for Critical Cyber-Physical Systems with a total fund of ~£130k. His current research interests include Cybersecurity, Wireless Communications, and AI applications in the Internet of Things and the Internet of Vehicles.  

Title: Build Trust and Security in your AI Solutions​: Exploring resilience and secure by design Cyber Systems (ALI) and Networks and Cyber Security for enabling Connected Vehicles and EV Charging

Abstract: Build Trust and Security in your AI Solutions​: Exploring resilience and secure by design Cyber Systems 

The greater use of AI has called for increased trust and security in raw/synthetic data quality and model decisions, especially in regulated environments such as finance. We have therefore developed a platform called Trust Your Machine Learning Output (TYMLO), which looks to address a key issue in the sector, by helping to improve synthetic data quality and model decisions for AI solutions. 

Standard AI tools, developed and adopted by most organisations, are generally not fit for purpose in these regulated settings due to the fact that they do not outline how the AI algorithms derive results. This has led to a growing lack of trust and security in the AI-based decisions and predictions being made, due to poor raw/synthetic data quality and a lack of explainability. 

To fill this gap, TYMLO (web-based toolkits helping data scientists, AI, and ML developers to trust ML outcomes) was devised by an expert team at TYMLO Technology Ltd. Consisting of three toolkits, TYMLO ensures trust, security, quality, and explainability to AI-based decisions. It’s designed to help reduce unreliability, improve raw/synthetic data quality, and add trustworthiness to any AI platform. 

This talk will also cover the aspects of resilience and security in the designing face of cyber systems.