Conference Agenda

Generative AI tools are fundamentally changing how students engage with engineering education by removing traditional barriers to obtaining solutions. This paper examines an undergraduate control engineering course redesigned to explore learning dynamics under reduced structure and unrestricted access to AI tools. Using a control-theoretic perspective, learning is interpreted as a dynamical system. An initial open-ended design revealed fragmented, event-driven learning behavior, with a primary bottleneck in structuring and integrating knowledge rather than accessing solutions. In response, structural and feedback mechanisms were introduced during the course, including project-based learning, iterative workflows, and peer review. These interventions can be interpreted as reintroducing feedback into the system, shifting it from an unstable open-loop regime toward a more stable closed-loop configuration. The results suggest that in AI-supported environments, the primary role of engineering education shifts from delivering knowledge to stabilizing learning processes through structure and feedback.

In today's digital society, driven by Artificial Intelligence (AI), Industry 4.0, and the Internet of Things, increasing internet connectivity makes web applications and data more vulnerable. Security breaches at large companies, affecting millions of users, are becoming more common. The rise of AI and growing curiosity about its use are making users more careless, which opens the door to hacker attacks. The human factor is a primary cause of vulnerabilities and security breaches, so ongoing and timely training is crucial for prevention. The article highlights the main cyber threats and vulnerabilities, as well as current testing and prevention tools, and emphasizes the importance of cybersecurity training for students, not only in IT majors but also in other fields. These skills are essential for their future success in the AI era. To highlight the significance of this issue, a simulated attack is performed on a test web application, and its potential impact is examined. For demonstration, tools like BurpSuite and SQLMap are used, with the actual testing target being the OWASP Mutillidae II web application. A survey exploring habits and attitudes toward security when using web applications was also conducted. The main survey results indicate that most respondents are unfamiliar with common types of threats and cyberattacks and wish to learn more about them. Many of them reuse the same passwords and fail to verify whether a site is secure, indicating poor cyber hygiene and awareness. Finally, the future of web application security and the role of artificial intelligence (AI) in it are discussed.

This research proposes and designs the Polygenic-weLLbeing (PLL) robot, an embodied service robot platform for domestic settings involving vulnerable groups. The system is designed to explore the integration of cooking skills education and dietary guidance, health and care support, and home Internet-of-Things control within a unified framework, with the aim of supporting user profiling and potential cross-module task coordination.To conduct a preliminary, prototype-level validation, this research adopted a mixed-methods approach and carried out two exploratory trials. The first was qualitative descriptive research in an aged-care facility setting, and the second was an embedded questionnaire survey conducted during cooking activities in an ethnic minority community. The results suggest that, in the older adult care setting, the PLL prototype showed initial indications of acceptability in terms of reminders, companionship, step-by-step guidance, and structured support, with older users showing a preference for concise language, clear steps, and a moderate pace of interaction. In the ethnic minority community setting, matching cultural dietary preferences, clear task guidance, and easy-to-understand voice prompts were associated with improved perceived acceptability and usefulness. The findings suggest that the design of healthcare robots for vulnerable groups should go beyond simply adding functions, but must also give due consideration to interactional comprehensibility, cultural adaptability, accessibility, and emotional support. As prototype research, this paper conducted scenario-based preliminary validation of two key modules of PLL, but it has not yet systematically evaluated multi-module coordination or the operation of the system. Future work is still needed to further verify its multi-module collaborative capability and its potential for sustained application in longer-term, larger-scale real-world settings.

This study explores the integration of Minecraft as a game-based learning tool in teaching ”Programming Fundamentals” to non-technical university students over three academic years. By analysing student performance and engagement levels, the research demonstrates significant improvements in programming comprehension, problem-solving skills, and academic success. The results highlight increased average marks, higher pass rates, and decreased awarding gap between White and BAME students, as well as a rise in good honours for both White and BAME students. Furthermore, entry-level qualifications did not hinder performance, suggesting Minecraft’s inclusivity across diverse student backgrounds. The findings highlight the potential of game-based learning and assessment to enhance engagement and create a more accessible learning environment in programming