Engineering mechanics is a fundamental part of many undergraduate programs, but it can also be one of the biggest hurdles at the beginning. This is due to the difficult content but also due to the challenge of conveying the practical relevance to students. The further development of the introductory courses in engineering mechanics has two objectives. The first objective is to offer students a more motivating approach to this challenging subject. This involves working on more practical examples by leaving the classroom and conducting experiments or shifting the focus from tedious calculations with pen and paper to experimenting with solutions. The second objective is to prepare students for the future. Aspects of programming and AI can be covered in separate courses in higher semesters. However, it is more efficient and sustainable to achieve this goal if these topics are already covered in introductory courses like mechanics from the first year onwards. This allows students to become true digital natives over the course of their undergraduate program. This publication takes a closer look at three technologies: virtual reality (VR) applications, the use of smartphones as measuring devices, and the use of numerical algorithms and the Julia programming language in the age of GenAI-Tools. It shows how these three technologies can contribute to the above-mentioned objectives in courses on engineering mechanics.
Higher education academic scheduling is a complex administrative undertaking with a direct impact on the efficiency of institutions, resource utilization, and the overall student experience. The development of efficient timetables is a process of coordinating various moving components, including faculty availability, classroom size, student requirements, and institutional regulations, all of which must be considered simultaneously. The paper discusses how the Arab Open University (AOU) addressed these issues by developing and implementing a rule-based, automated timetabling artificial intelligence (AI) solution. AOU’s multinational and diversified student body of around 62,000 students and more than 1,000 staff members within its nine regional branches makes a manual approach to scheduling slow and prone to errors. The paper presents the transformation of AOU to a modern Python-based AI solution as an example of transitioning from traditional labor-intensive scheduling methods to a more modern approach. It outlines the methodology, system structure, challenges encountered during the implementation process, and quantifiable benefits. The rule-based approach was more transparent, flexible, and aligned with institutional priorities than black-box optimization algorithms like genetic algorithms or simulated annealing. The findings were impressive: the time required to create timetables was reduced by four to six weeks to less than two hours, the number of conflicts in the schedule decreased by 85%, and the use of the classroom increased by 65% to 78%. These results suggest that university timetabling can be modernized using clear, rule-based AI systems that do not compromise oversight and trust within the university.
The implementation of online assessment in Namibian higher education institutions (HEIs) is challenged by limited infrastructure, differential digital literacy levels, and disconnection from pedagogical frameworks, mainly in open and distance learning (ODL). This work discusses a phased model for online assessment that addresses these challenges while also promoting inclusivity, scalability, and academic integrity. A qualitative method was used to collect data from students, educators, and administrators from four institutions in Namibia. The findings imply that the deployment of online assessments ultimately depends on institutional and stakeholder readiness, sound infrastructure, and assessment design rather than on technology adoption alone. The framework identified comprises planning, platform configuration, training sessions, pilot testing, implementation, monitoring, and ongoing review. Based on theories such as the Diffusion of Innovations, the Technology Acceptance Model, and Cognitive Load Theory, this model provides context-specific implications for a high-tech-orientated perspective on higher education systems in the transition to digital assessment environments. This study addresses a widely unmet need in the implementation of online assessment in developing contexts and demonstrates a model to provide a suitable solution to advance equity, quality assurance, and capacity in online assessment practices in institutions in developing countries.
The study developed and applied a computational algorithm that selects the best medical images, aiming to improve pattern recognition skills among medical students and professionals. The algorithm, written in MATLAB, uses six quantitative metrics—sharpness, contrast, brightness, entropy, processing time, and structural artifacts—and is applied to physiological or pathological magnetic resonance neuroimaging to identify the image that most contributes to improving the visual and diagnostic performance of groups of participants. This is achieved through combinatorial weight analysis (56 combinations), in which the image with the highest number of wins is selected. The system was applied to a study with 60 participants divided into three groups (medical students, residents, and specialists in neurology), who answered two questionnaires: the first consisting of raw images selected at random, and the second using images previously processed and selected by the algorithm. The data were statistically analyzed, showing improvement in absolute and relative performance when comparing input and output, with emphasis on the domains of topographic anatomy and zonal recognition of structures. The results indicate that the use of images optimized by objective criteria improves the accuracy of image identification and participant confidence. It is concluded that the use of reproducible and measurable computational solutions is relevant in medical qualification, improving educational and diagnostic performance.
Teaching programming language syntax across multiple languages presents a significant cognitive challenge for undergraduate students, compounded by the absence of inclusive pedagogical strategies for deaf learners. Reported here is an action research study conducted at Universidad Politécnica de Santa Rosa Jáuregui (UPSRJ), Querétaro, México, in which a didactic strategy mediated by an inclusive virtual world was co-designed and implemented with programming instructors and students at the TRAMVET Laboratory. Developed on the Sansar platform, the virtual environment integrates Mexican Sign Language (LSM) representations through three-dimensional avatar animations, Universal Design for Learning principles in spatial and iconographic organization, and an immersive syntax-differentiation activity that provides immediate visual feedback. Grounded in the sociocritical paradigm and action research methodology, the investigation followed iterative cycles of diagnosis, planning, action, observation, and reflection. Combining multimodal affordances of 3D immersive digital environments with deliberately inclusive spatial design strengthens students’ capacity to distinguish syntactic structures across Python, Java, C, and JavaScript, while simultaneously reducing communicative barriers for deaf participants. Contributing an empirically grounded didactic model, this paper identifies design principles transferable to comparable institutional contexts.
This study examines the growth and intellectual structure of scholarly literature on responsible generative artificial intelligence (GAI) in pedagogy and proposes a context-sensitive conceptual framework for responsible GAI adoption in the Global South. Using bibliometric analysis and systematic review, the study analyzes publications published between 2022 and 2025 extracted from the Dimensions AI database. The bibliometric findings indicate a rapid expansion of the field, with an annual publication growth rate of 18.92%, reflecting increasing academic interest in GAI-enabled pedagogy. Medical education journals dominate the research landscape, with BMC Medical Education (22 articles; 227 citations) and JMIR Medical Education (13 articles; 355 citations) emerging as the most productive sources, while Scientific Reports exhibits the highest citation impact (8 articles; 628 citations). In terms of geographical contribution, the United States (40 articles), China (35 articles), and Australia (16 articles) lead in research output, whereas selected Global South countries, particularly the United Arab Emirates (497 citations), demonstrate high scholarly influence despite lower publication volume. Thematic analysis reveals a strong focus on artificial intelligence, generative AI, medical education, and language teaching, alongside emerging but underdeveloped themes related to ethics, governance, and sustainability. Insights from the systematic review identify ethical awareness, trust, institutional support, infrastructural readiness, and AI competence as key determinants shaping adoption. Accordingly, synthesizing these findings, the study advances an extended TAM that conceptualizes responsible GAI adoption within pedagogical systems of the Global South.
This study is set out to examine the current situation of artificial intelligence (AI) application for Hanoi Metropolitan University’s faculty members; review challenges and opportunities; and propose a strategic framework of AI integration for teaching and learning. A mixed approach was employed, involving a survey with 156 lecturers across disciplines and sources of qualitative data—institutional documents and in-depth interviews with 20 lecturers. Theoretical background the study is based on the theoretical foundation of the technology acceptance model (TAM). The results suggest that a majority of faculty (67.9%) have tried out AI tools due to the hype of surface-level AI such as generative AI (e.g., ChatGPT) but are not integrating it into their educational practices, with 12.8% using it daily. There were major barriers such as absence of formal training (only 15.4% had any institutional training), lack of infrastructure, and perceived risks to academic integrity and data quality. There was substantial interest in professional development, with a request from 89.1% of faculty members for the AI-related courses. The study, in turn, reveals a real difference between what AI can do and what is used at the Hanoi Metropolitan University (HNMU) for teaching purposes. It argues that a strategic, top-down approach is required to transition from ad hoc use of digital technologies through to meaningful integration. A phased deployment model is recommended, with an emphasis on faculty development, infrastructure support, and well-defined institutional policy. These findings have implications for HNMU and other Vietnamese HEIs that are struggling with the complexities of digital transformation.
This study aims to provide a comprehensive assessment of the current state and near-term trajectory of augmented and immersive ecosystems, synthesizing progress across key dimensions of interoperability, identity, and hardware innovation. The analysis extends the “Augmentiverse” framework—an Augmented Reality (AR)-first, standards-aligned precursor to a broader Metaverse—by integrating tangible, real-world developments in AR smart glasses and head-mounted displays. It further situates these developments within the emerging paradigm of spatial computing, where digital content is persistently anchored to physical environments through shared standards and environmental understanding. The study examines how cross-industry coordination across XR runtimes (OpenXR, WebXR), 3D asset/scene interchange (glTF, OpenUSD via AOUSD), decentralized identity (W3C DIDs and Verifiable Credentials 2.0), immersive media (OMAF), and future network capabilities (IMT-2030) provides the essential foundation for a persistent, interoperable digital layer. Evidence from major technology firms including Meta’s strategic hardware ladder (Ray-Ban Meta Glasses, Oakley Meta HSTN, Meta Ray-Ban Display Glasses, and Orion) and the broader competitive landscape featuring Apple, Samsung, and Microsoft demonstrates that a pragmatic, staged approach is the most viable path to population-scale adoption. The central contribution of this research is a detailed, standards-first roadmap that operationalizes this vision, explaining how the convergence of lightweight, socially acceptable eyewear with robust open standards reduces fragmentation, strengthens user trust, and creates immediate economic value, thereby paving a realistic path toward deeper immersion as foundational technologies mature.
This systematic review critically examines the growing field of artificial intelligence (AI) applications in tracking student engagement and disengagement in educational settings. We synthesize current literature, employing bibliometric analysis to understand the complexities of technology-integrated teaching methods and their effectiveness in creating engaging learning environments. This study employs a rigorous methodological framework, incorporating the preferred reporting items for systematic reviews and meta-analyses (PRISMA) model and the population, intervention, comparison, outcomes, and study design (PICOS) criteria to ensure a structured and comprehensive review. A systematic search strategy was implemented to identify relevant studies from authoritative academic databases. The research findings indicate a significant use of new datasets and virtual learning environments, particularly emphasizing higher education. Despite the promising advancements in AI-driven engagement detection, our analysis reveals critical research gaps, such as the lack of detailed demographic information, especially the age factor that greatly influences engagement behaviors. This absence highlights the need for more specific engagement detection tools suitable for different educational levels. Another key observation is the limited research on early education, a critical area where engagement is crucial yet subtly indicated. Considering these points, we offer recommendations for future research, calling for a comprehensive approach that includes detailed demographics, integration of various learning settings, ensuring broad technology access, improving multimodal techniques, and maintaining privacy and ethical standards. The study’s practical implications underscore the need for more adaptable, inclusive, and ethically responsible technological contributions to education, benefiting educators, policymakers, and AI developers.
In the game-based learning (GBL) literature, it is widely accepted that digital educational escape games (DEEG) contribute to increased academic performance, improved motivation, and increase in learner engagement. However, while these results are empirically well documented, there is a lack of insights into the cognitive and neural processes behind the reported results. This pilot study attempts to address this issue through an interdisciplinary research design. It includes both GBL and educational neuroscience (EN) to understand which mental processes are involved when interacting with a DEEG. The study engaged 23 adults, divided into three experimental groups. During their interactions in the different experimental conditions, changes in their brain waves were recorded via the use of an electroencephalogram (EEG). Specific attention was given to the neural markers of memory encoding/retrieval (hippocampal theta waves) as well as concentration (low beta waves) and the participants’ perceived usefulness of the activity (frontal alpha wave asymmetry). The study showed a significant increase in memory encoding/retrieval among the participants interacting with the DEEG. This increase was found to be linked to the participants’ perceived usefulness of the activity.
This study aims to explore Portuguese teachers’ perceptions of their digital teaching skills, particularly regarding Web 2.0 tools for online instruction and learning. A total of 97 teachers participated in this quantitative research, responding to a questionnaire comprising 38 questions, including open-ended items for both in-service and pre-service teachers. The findings highlight the need to further develop digital competencies, especially in selecting and using open educational resources (OER) effectively. Teachers recognised the value of digital tools in enhancing both teaching quality and student learning outcomes. The results also underscore the importance of equipping educators with the skills to evaluate and apply relevant digital materials in their practice. Micro-credentials emerged as a strong motivator, with the majority of participants expressing interest in earning certification through professional development opportunities. This reflects a broader desire among educators to improve their digital expertise and advance their careers. The study offers both theoretical and practical insights for educational institutions, suggesting that investment in digital skill development and certification pathways can significantly enhance teaching effectiveness and learner engagement in digital contexts.
This pilot study explores the feasibility of integrating large language model (LLM) assistants into physics education through a Moodle plugin designed to address conceptual understanding in Newtonian mechanics. Using OpenAI’s GPT for real-time Socratic dialogue, the plugin guides students through misconception-targeted questions adapted from the Force Concept Inventory (FCI). Aligned with principles of inquiry-based learning, the intervention compares AI-guided feedback with instructor-guided materials over a one-week, three-session classroom study. Results suggest that students receiving AI-guided Socratic dialogue showed greater conceptual gains on certain targeted items (e.g., Newton’s Third Law), whereas instructor guidance proved more effective for other concepts (e.g., mass and free-fall independence). Survey feedback highlights the immediacy and interactive nature of the AI while also noting a preference for the clarity provided by instructors. Qualitative analysis of open-ended question responses suggests that AI-driven dialogue promotes deeper reasoning when restating student ideas and scaffolding reflection. These preliminary findings underscore the potential of LLMs to support conceptual change in physics education when thoughtfully embedded within learning management systems, highlighting the complexity and value of personalized, interactive feedback for addressing student misconceptions.
This study explores the integration of neural networks onto resource-constrained microcontrollers in academic teaching and learning settings in order to enhance students’ understanding of artificial intelligence. By leveraging MicroPython, the open-source framework AI-ANNE: (A) (N)eural (N)et for (E)xploration is used to facilitate the transfer of pre-trained models from high-level artificial intelligence libraries like TensorFlow and Keras to microcontrollers. A comparative analysis was conducted to assess students’ comprehension of neural networks under different instructional methods. Statistical evaluations using Tukey’s HSD post-hoc test revealed that students who implemented neural networks from scratch in MicroPython demonstrated significantly higher learning outcomes compared to those using TensorFlow and Keras in Python. However, no significant difference was observed between students who implemented MicroPython models on standard computing environments and those who deployed them onto microcontrollers. This suggests that explicit implementation in MicroPython fosters a deeper conceptual understanding of neural networks. While high-level artificial intelligence libraries like TensorFlow and Keras provide efficiency, they may obscure fundamental learning processes when it comes to trust, transparency, understandability, and usability as key concepts of explainable artificial intelligence approaches. Thus, direct implementation in MicroPython enhances comprehension and prepares students for responsible artificial intelligence development.
Cloud computing is a critical skill in today’s world, and nearly all universities worldwide offer online platforms to meet the high demand for computing courses. While these courses provide knowledge and hands-on practice, there remains a gap in instructional course design. This paper explores how online cloud computing courses are structured and to what extent they apply educational theories such as Bloom’s Taxonomy and Gagné’s Nine Events of Instruction. Using a qualitative, multiple case study approach, the paper analyzes three realworld examples: AWS Academy Cloud Foundations, Google Cloud Skills Boost, and a Scratchbased course developed at the University of Salerno. These cases show different approaches to managing content, guiding student progress, and supporting active learning. While some platforms focus on tools and task completion, others emphasize learning goals, feedback, and scaffolding. The study suggests that cloud platforms provide adaptability but often lack effective instructional design. Courses should focus on structured learning paths, incorporating reflection and feedback. This paper proposes a five-step framework to improve online cloud computing courses by integrating academic theory with practical delivery, contributing to digital education, and offering guidance for instructors, course designers, and institutions to enhance student engagement and skill mastery.
This study addressed the need to understand the process of technology adoption among rural schoolteachers despite the digital divide that influences their willingness to adopt emerging technologies as to what needs to be done to enhance teaching effectiveness. This understanding helps gauge the adoption process and identify the Level of Use (LoU) pertaining to technology adoption. The purpose of this study is to investigate the LoU of emerging technology adoption among rural schoolteachers in Sarawak and to what extent their LoU influences them to adopt emerging technologies. This study utilized LoU as one of the diagnostic dimensions of the Concerns-Based Adoption Model (CBAM). This study employed a qualitative research methodology using a multi-method of data collection, including in-depth interviews and document analysis. The result from the interview concluded that rural school teachers were in LoU I – Orientation, LoU III – Mechanical Use, and LoU V – Integration. In the analysis of interview data, three themes emerged pertaining to the extent to which rural school teachers’ LoU influences their willingness to adopt emerging technologies in their classroom practice: (i) teachers’ collaborative effort; (ii) learning experience; and (iii) professional development. This study can be regarded as a paradigm for rural schools seeking to adopt emerging technologies to augment instructional efficacy and improve pedagogical outcomes.
This systematic review study synthesizes and presents research findings relevant to the instructional application of gamified augmented reality (AR) tools in language learning from 25 peer-reviewed research from 2019 to 2023 with preferred reporting items for systematic reviews and meta-analyses (PRISMA) guidelines. The review purpose was to identify: (1) the commonly adopted AR games in language classrooms; and (2) students’ perceptions of learning languages with AR games. The main study findings suggest the focus of the past research efforts was placed mostly on self-designed AR games while there is a relative lack of research interest in premade or existing games for language learning. In terms of students’ perceptions, the findings indicate that students generally have positive perceptions toward AR games in the language learning experience. Moreover, games that feature a combination of 3D models and animations, including peer collaboration and competition, seem to be the favorite for digital native students. The findings also reported concerns from the students such as the need for more assistance from the instructors. Suggestions for modification of instructional design and future research in the field of AR-integrated language learning are also provided.
The education community continuously develops learner profile (LP) models to support decision-making in learning analytics (LA). However, a gap persists in aligning abstract stakeholder requirements with complex machine learning (ML) patterns. Educational stakeholders such as decision-makers, educators, and pedagogical engineers perceive and categorize LPs (e.g., learners in difficulty, active/inactive learners, those in progress, and success-oriented vs. at-risk learners) through mental models. These latter reflect real-world perceptions and pedagogical practices grounded in common educational concepts. To bridge this gap, data scientists must ensure technical ML insights align with stakeholder needs by selecting relevant features, addressing explainability, mitigating biases, and validating patterns against domain assumptions. For example, a learner generating extensive log data through repeated solution attempts may appear engaged from a human perspective but exhibit disengagement based on unexpected ML discovery patterns, highlighting biases in data interpretation or human perception. We propose Req2XAI (From Requirements to Explainable Machine Learning Models), a framework that establishes a bidirectional mapping between stakeholder requirements for LP analysis and ML-driven learner profiles. Req2XAI externalizes stakeholders’ mental model about LP via a conceptual model into requirements and goals and formalizes an end-to-end workflow, from stakeholder objectives to explainable ML models, ensuring transparency at each stage. A proof-of-concept prototype is implemented through a use case, considering the requirements of the Steering Committee of the écri+ project. This work introduces open research challenges associated with the Req2XAI framework, which merit further exploration.
This study focused on the phenomenon of technological anxiety in the contemporary art world within the context of artificial intelligence (AI) technology. By analyzing cases of anti-AI movements initiated by artists on platforms such as ArtStation and GitHub, this study identified the key issues that arise in the art world under the influence of technological anxiety. A dialectical analysis of two central issues—intellectual property and occupational substitution—was conducted from four perspectives: technology, legal principles, user perspectives, and historical context. Furthermore, the study discussed core artistic values and compliance anxiety, particularly as they relate to the concerns of art students. Qualitative research methods, including content analysis and case studies, were employed. The study uncovered the underlying reasons for technological anxiety in the art world, critiqued the irrational aspects of the anti-AI movement, and offered solutions to alleviate technological anxiety. Additionally, it provided recommendations for the career development of practitioners. The study also highlighted the importance of core artistic values, the elimination of compliance anxiety, and the improvement of teachers’ proficiency in AI as key areas for enhancing student education.
One of the key areas of interest within learning analytics is identifying student similarities to support collaborative applications such as score prediction, personalized recommendations, and group formation. Clustering is a prominent method for grouping students based on shared learning behaviors, enhancing peer learning, and fostering communities in online courses. This study introduces an intuitive graphical clustering approach using activity logs that track student engagement with different learning resources. These interactions are modeled as multi-dimensional vectors, and a social network of learners is constructed using cosine similarity. Social network analysis (SNA) is then applied to detect learner communities. The dataset for implementation and evaluation includes activity logs and grades from 792 students in an undergraduate study program. Results indicate that learners in the same clusters have similar interaction patterns and grade point averages (GPAs). Statistical measures, such as silhouette index and root mean square standard deviation (RMSSTD), demonstrate the method’s effectiveness and benchmark its performance against K-means clustering. This approach shows significant potential for uncovering and visualizing implicit learner groups.
