Emerging Technologies in Cyber-Physical Systems: A Global Seminar Series

Presentation

Presentation

Presentation

Presentation

Presentation

Presentation

Presentation

Presentation

Presentation

Presentation

Presentation

Presentation

Title: Brain machine interfaces to control assistive devices

Abstract: Brain-machine interfaces are computational algorithms that interpret neural activity and convert it into signals to control assistive devices such as prostheses or exoskeletons. In this talk we will discuss the process of designing brain-machine interfaces, from the design of experiments to measure electroencephalographic and kinematic signals, to the implementation of algorithms to relate these signals and generate control signals for assistive devices.

Bio: Griselda Quiroz Compeán is an electronic engineer with graduate studies in control and dynamic systems at IPICYT-CONACYT, MX. Since 2009 she is a full time research professor at the School of Mechanical and Electrical Engineering (FIME) of the Universidad Autónoma de Nuevo León (UANL). She is part of the national system of researchers of Mexico. Her work has been reported in more than 20 scientific articles, more than 30 participations in international conferences, she has developed 10 funded research projects and has directed 18 theses at bachelor, master, and doctoral level. Her research interests are in mathematical modeling and control of physiological systems and their applications in brain-machine interface design, design and control of motor assistance devices and mathematical modeling and control of glucose in diabetes. Griselda is an active member of the PIIT Nuevo León Women IDi Committee, of the IEEE Monterrey Section and is a mentor for STEM women, certified by the British Council. In addition, she is currently the coordinator of special projects of the International Relations Secretariat of FIME-UANL.

Title: Design and evaluation of AI models and algorithms for generating personalized learning trajectories.

Abstract: Education research is seeking to revolutionize how students learn. Imagine a dynamic and personalized learning plan that adapts and evolves for each student, like a roadmap to student success. In this talk, we will present a conceptual framework based on Artificial Intelligence techniques and Mathematical Programming to model curriculum information in order to generate personalized learning trajectories. We will explore how these approaches can adapt to subjective factors in the educational process, such as student stress and learning effects. Additionally, we will discuss open research problems and opportunities to develop personalized learning paths to enhance learning in autonomous and digital environments.

Bio: Romeo Sanchez is a professor affiliated with the School of Mechanical and Electrical Engineering (FIME) at the Autonomous University of Nuevo Leon (UANL), Mexico. Romeo holds a Ph.D. in Computer Science from Arizona State University. He is the leader of the Computational Intelligence research group and the academic coordinator of the Artificial Intelligence Engineering program. His research interests include automated planning and scheduling, agent optimization and heuristic search control, intelligent decision support, and distributed systems engineering. His most recent work has focused on developing planning models for education and transportation domains and applying classification algorithms to medical problems.

Title: Formation flight of multiple UAVs

Abstract: The robust consensus of the multiple quadrotors for formation flight is proposed as a solution for the multiagent system (MAS) problem. The Newton-Euler formulation is used in order to describe the mathematical model of the N quadrotors considered as agents and a nonlinear sliding mode control stabilizes the translational and rotational dynamics of each agent so that this control algorithm drives the general sliding manifold to zero in finite time. This general sliding manifold consists of a sliding surface for the navigation of each agent and an auxiliary sliding surface for the consensus of the MAS. Then, the robust consensus algorithm guarantees that the MAS executes the formation flight and pursuit in the trajectory tracking even in presence of disturbances. Finally, real-time experiments show that the MAS successfully reaches the consensus.

Bio: Octavio Garcia Salazar obtained a B.Sc. in electronic engineering and an M.Sc. in electrical engineering with specialization in robotics at the Technological Institute of La Laguna, Torreon Coahuila, Mexico, in 2000 and 2003, respectively. He obtained a Ph.D. in control systems at the University of Technology of Compiègne, France, in 2009. From January 2010 to December 2011, he held a post as a CNRS postdoctoral researcher in autonomous systems at the laboratory LAFMIA UMI 3175 CNRS-CINVESTAV Mexico. From January 2012 to December 2012, he was a visiting researcher at CINVESTAV Monterrey. Since January 2013, he is professor-researcher at the Faculty of Mechanical and Electrical Engineering of the Autonomous University of Nuevo Leon, San Nicolas de los Garza Nuevo Leon, Mexico. His research interests are guidance, navigation, and control of UAVs, UAS, flight dynamics, avionics, and robotics.

Title: Protecting IoT Obsolete Devices in Critical Infrastructures – H2020 CyberTrust Project

Abstract: The H2020 Cyber-Trust project was an ambitious European initiative focused on enhancing cybersecurity for the Internet of Things (IoT) ecosystems. The project, which involves a consortium of partners from various sectors, aimed to create a robust framework for the detection, analysis, and mitigation of cyber threats targeting IoT devices and networks. It leveraged cutting-edge technologies such as blockchain, artificial intelligence, and advanced data analytics to provide comprehensive cybersecurity solutions. The url of the project can be found here https://cyber-trust.eu/ A significant aspect of the Cyber-Trust project is the development and implementation of an advanced Intrusion Detection System (IDS). The IDS within the Cyber-Trust framework is designed to address the unique challenges posed by IoT environments, which include diverse device types, varying communication protocols, and resource constraints. This presentation is focused on the IDS research work done and we will explore how machine learning was utilised along with blockchain and graph theory in order to create a novel solution to mitigate zero-day attacks. 

Bio: Dr. Stavros Shiaeles (male) is Assoc. Prof. in Cybersecurity, at University of Portsmouth. He has extensive research experience that spans several critical areas within the cybersecurity domain. His work has notably contributed to the advancement of knowledge in network security, malware detection, and the development of robust cybersecurity measures against advanced persistent threats (APTs). Dr. Shiaeles has a strong publication record, featuring in respected journals and conferences, underscoring his commitment to advancing the state of cybersecurity through rigorous research. In addition to his academic prowess, Dr. Shiaeles holds several professional certifications such as CEH, CAST611, CCNSP and COBIT5, that attest to his expertise and practical skills in cybersecurity and demonstrate his comprehensive understanding of both the theoretical and practical aspects of securing digital infrastructures against increasingly sophisticated cyber threats. 

Title: Fractional calclulus in modern control systems

Abstract: Industry 5.0 is considered a new production model where high performance expectations must be fulfilled. It is not about classical control systems, it is about the control of cyber-physical systems. It demands more efficient controllers with high dynamic performance. Many real dynamic systems are better characterized using a non-integer order dynamic model based on fractional calculus or, differentiation or integration of non-integer order. Traditional calculus is based on integer order differentiation and integration. The concept of fractional calculus has tremendous potential to change the way we see, model, and control the nature around us. In this lecture, it is offered a tutorial on fractional calculus in modern control systems. Basic definitions of fractional calculus, fractional order dynamic systems and controls are presented first. Then, fractional order PID controllers are introduced which may make fractional order controllers ubiquitous in industry.

Bio: Eva H. DULF (Senior Member, IEEE) received her Ph.D. degree from the Technical University of Cluj-Napoca, Cluj-Napoca, Romania, in 2006, where she is currently Professor in the Automation Department. She is also research professor at Obuda University, Hungary. She has published more than 150 papers and received 43 awards at prestigious International Exhibitions of Inventions. Her research interests include modern control strategies, fractional order control, modelling of biochemical and medical processes.

Title: Applications of AI in Ambient Assisted Living

Abstract:  Ambient Assisted Living (AAL) is concerned with using various technological solutions to allow people with additional care needs to live independently in their preferred environment.   The health and social services are struggling and the pressure on the statutory services will continue increasing as the population ages. Therefore, solutions that can help manage more conditions in the home, reduce hospital stays, and promote wellbeing will be crucial for addressing the impending care crisis. Previous AAL work concentrated mainly on ensuring safety in the home by monitoring for well-defined emergencies, such as falls, or helping users with a specific task such as medication compliance. To meet future care needs, the next generation of AAL will need to be systems of interconnected devices that support the lifestyle of the person in their own home and help carers prioritise the interventions that they make. The focus of our research is on monitoring the performance of Activities of Daily Living (ADLs) and ensuring that they are being undertaken to promote wellbeing as well as safety. We explore applications of Artificial Intelligence (AI) for recognising and tracking of ADLs. In addition to considering technical implementation aspects of this problem we also investigate the issues of technology adoption by users.

Speaker bio:  Dr Rinat Khusainov is Associate Head of Computing (Research and Innovation) at the University of Portsmouth. He also leads and manages University’s Cisco Networking Academy. He holds PhD from University College Dublin and has more than 20 years of experience in applied AI, embedded systems, and computer networks, with applications in health technologies and wellbeing, information management, computer and network security. He worked on projects funded by the Irish and UK governments, and the European Commission, including knowledge transfer partnerships with Irish and UK SMEs in data communications and care sectors. He published over 60 refereed articles and served as local organising chair and program committee for several international conferences. He is Member of the IET, Fellow of the HEA, and Chartered Engineer.

Title: Privacy-preserving AI tools to respond to and understand people

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.

Bio: 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.