Computer Applications in Engineering Education最新文献

This study examines the impact of Napkin AI, a generative artificial intelligence tool, on student learning outcomes and knowledge retention via creative mind mapping. The research model, which combines the Unified Theory of Acceptance and Use of Technology 2 (UTAUT2) with the Information Adoption Model (IAM), comprises six principal independent variables: Social Influence, Learning Support Environment, Ease of Use, Perceived Benefit, Information Quality of Mind Mapping, and Usefulness of Mind Mapping. Mediating variables comprise Behavioral Intention to Use, Actual Use of Mind Mapping, and Cognitive Engagement in Mind Mapping, while Knowledge Retention and Learning Outcome serve as dependent variables. The study employed a quantitative approach with a sample of 570 Vietnamese university students majoring in STEM. Data analysis was conducted using Structural Equation Modeling (SEM) with IBM SPSS 25 and AMOS 24. The model exhibited a satisfactory fit (χ²/df = 1.906, GFI = 0.889, CFI = 0.935, TLI = 0.929, RMSEA = 0.040, PCLOSE = 1.000), with all direct routes being statistically significant (p < 0.001 to p = 0.037). The findings underscore the significance of user perception and involvement in improving the educational efficacy of AI-driven mind mapping tools, providing practical implications for technology integration in higher education.

This study presents the design and implementation of a low-cost 3-degree-of-freedom (3-DOF) helicopter platform aimed at strengthening control education through the integration of digital twins (DTs) and hardware-in-the-loop (HIL) systems in a Rapid Control Prototyping (RCP) framework. The methodology consisted of first modeling and understanding the nonlinear dynamics of the system, and then building the physical platform using a Raspberry Pi Zero W and a 3D-printed structure, enabling both affordability and a carry-it-home (CIH) approach. Once constructed, the platform was calibrated to match its DT, after which pole placement and Linear-Quadratic-Gaussian (LQG) controllers were designed, simulated, and tested on both the DT and the HIL setup. To assess the educational impact, pre- and post-course questionnaires were applied to gather students' perceptions and expectations. The results indicate that the experience not only improved comprehension of feedback control concepts but also provided a more comprehensive and practical understanding of DT and RCP technologies.

Discrete Mathematics is an important and challenging course for computer science and engineering students. It includes topics, such as logic, sets, proofs, number theory, graphs, trees, computation, relations, functions, and basic algorithmic concepts. These topics require strong analytical reasoning and consistent effort. As a result, many students find this course challenging to perform well. The aim of this study is to predict student performance in a Discrete Mathematics course at a reputed private university located in Bangladesh. Data were collected from both course instructors and students during the spring and summer semester of 2025. Instructors provided academic records, such as attendance, quizzes, assignments, and midterm scores. Students provided additional information, which included daily study time, subject interests, and use of learning platforms. The final data set included records for 240 students. K-means clustering with the Davies–Bouldin method was used to group similar students. Then, four machine learning (ML) models were trained and tested: Support Vector Machine (SVM), Decision Tree, K-Nearest Neighbors, and Naïve Bayes. The models were implemented using Python's scikit-learn library, with stratified sampling and fivefold cross-validation. Among the models, SVM achieved the highest accuracy of 96% after parameter tuning. Naïve Bayes had the lowest accuracy due to the assumption of feature independence. Key predictors of performance included mean score, attendance, and daily study hours. Findings show that ML can help instructors identify at-risk students early, provide focused academic support, and improve learning outcomes. While the results are promising, the study is limited by sample size and does not include psychological or emotional factors. Future work will explore larger data sets and apply interpretable Artificial Intelligence techniques for better model transparency.

Process Systems Engineering (PSE) is a pillar of Chemical Engineering. It has undergone a great change during the last decades due to the depletion of natural resources, environmental regulations, and increased global business competitiveness. All this has led to an evolution towards the integration and intensification of processes, which requires solving complex and multivariate decision problems. Process optimization provides a scientifically proven methodology for finding the best alternatives in making these decisions through mathematical and computational models. This paper describes a set of exercises integrated into a project that aimed to teach basic concepts of multivariable optimization in the real context of the refining and petrochemical industry within a course in PSE in the BSc degree in Chemical Engineering. For this, the proposed cases were solved using MS Excel Solver. The experience was successful since the students were able to learn the concepts and acquire the skills needed to solve the cases successfully, and this was reflected in the grades obtained. Furthermore, the survey results on the activity, employed to get feedback from the students, showed excellent acceptance.

Expert systems are a core branch of Artificial Intelligence (AI), but they are often taught through abstract theory that hinders students' understanding and motivation. This study presents BattleInTheSky, a gamified expert system implemented on the low-cost Raspberry Pi Pico W microcontroller using MicroPython. Designed as a hands-on learning tool for Mechatronics Engineering students, the system uses IF–THEN rules to control hardware components including buttons, an OLED display, and a buzzer. It stores gameplay statistics locally and optionally uploads data remotely, providing an interactive experience that connects AI theory to practical applications. The project integrates symbolic AI, embedded programming, gamification, and principals of ethical design. It was implemented through a constructivist learning sequence in a sixth-semester Expert System course, guiding students from basic hardware control to the design of fully functional expert systems. A mixed-methods evaluation assessed technical performance and educational impact. Results showed that students improved their understanding of expert systems by 22.4% (Cohen's d = 1.3) and reported high motivation (mean 4.0/5). Technical evaluation confirmed the system's reliability, responsiveness, and usability in classroom settings. The project also fostered ethical reflection through rule transparency and responsible decision-making logic. BattleInTheSky demonstrates how low-cost, transparent, and rule-based AI systems can promote accessible and engaging AI education. The approach illustrates how transparent, rule-based systems can support explainable and accessible engineering and computing courses, particularly in resource-constrained contexts.

Currently, there is a mismatch between traditional Linux teaching and dynamic learning needs. Therefore, this study proposes a dual drive intelligent teaching model based on constructivism and game theory, and constructs a collaborative optimization framework of “task chain-resource game-container scheduling”. Firstly, design a cognitive transition task chain represented by weighted directed graphs, dynamically adjust node weights through reinforcement learning, and reduce the standard deviation of high-order task completion time; Secondly, establish a tripartite game engine and use multiagent Q-learning to solve Nash equilibrium; Finally, develop a lightweight Docker cluster and voice interaction system to reduce container response latency through a dual objective optimization algorithm. Empirical research has shown that the experimental group achieved a score range of 90+in the Linux certification exam, accounting for 36%. The code quality Abstract Syntax Tree (AST) similarity increased by 19.2%, and the system maintained a task completion rate of 92.7% in high concurrency scenarios. This model provides a replicable and scalable paradigm for current intelligent teaching.

As generative artificial intelligence (GAI) continues to advance, its integration into engineering education offers new opportunities to enhance student's engagement and comprehension in complex subjects. This study proposes a GAI-supported BOPPPS instructional model (G-BOPPPS), which embeds GAI tools into six stages of the traditional BOPPPS to foster active learning. The model incorporates adaptive goal setting, automated assessment generation, interactive learning activities, and data-driven feedback mechanisms. A quasi-experimental study was conducted in an undergraduate Operating System course involving cohorts from 2020 to 2022. The results indicate that the G-BOPPPS model significantly improved the academic performance of the students. A post-course survey revealed that more than 90% of the students perceived GAI to be beneficial for enhancing participation, conceptual understanding, and self-directed learning. The follow-up interviews further demonstrated that GAI integration facilitated the shift of the learning mechanism from passive knowledge reception to active knowledge construction. This study provides an empirical framework for embedding GAI within student-centered instructional models, which is applicable to the education of computer science and other engineering disciplines.