Computer Applications in Engineering Education最新文献

Machine learning (ML) has the potential to enhance educational predictive analytics, but its adoption is limited by the programming expertise required to develop models. Traditional ML tools require coding skills, which makes them inaccessible to educators and researchers without computational backgrounds. Existing no-code platforms lack affordability and accessibility. This study addresses this gap by developing and validating a no-code ML builder to enable non-programmers to build, evaluate, and deploy ML models. Design and development research approach was adopted in the study. It utilizes Python-based tools such as Streamlit and scikit-learn. The tool underwent expert validation and comparative performance testing against Google Colab using datasets from Kaggle, consisting of 5000 and 2392 student performance records. The results show that the no-code ML builder, which is accessible at nextml.streamlit.app achieved a predictive performance comparable to coded models. A minor performance gap was observed in some algorithms, with Logistic Regression achieving an accuracy of 63.88% compared to 73.28% in Google Colab. Experts in educational technology and computer science rated the tool highly for usability, with mean scores ranging from 4.33 to 4.57. 71% of evaluators found it suitable for educational datasets, and 56% endorsed its ability to handle students' data sets. The study concludes that the tool bridges the accessibility gap in the application of ML in education while maintaining competitive model performance. It recommends that Institutions adopt no-code tools. Future research should focus on incorporating more complex algorithms.

Large Language Models (LLMs) are transforming several aspects of our lives, including text and code generation. Their potential as “copilots” in computer programming is significant, yet their effective use is not straightforward. Even experts may have to generate multiple prompts before getting the desired output, and the code generated may contain bugs that are difficult for novice programmers to identify and fix. Although some prompting methods have been shown to be effective, the primary approach still involves a trial-and-error process. This study explores mechanical engineering students' experiences after engaging with ChatGPT to generate code for the Finite Element Analysis (FEA) course, aiming to provide insights into integrating LLMs into engineering education. The course included a scaffolded progression for students to develop an understanding of MATLAB programming and the implementation of FEA algorithms. After that, the students engaged with ChatGPT to automatically generate a similar code and reflected on their experiences of using this tool. We designed this activity guided by the productive failure framework: since LLMs do not necessarily produce correct code from a single prompt, students would need to use these failures to give feedback, potentially increasing their own understanding of MATLAB coding and FEA. The results suggest that while students find ChatGPT useful for efficient code generation, they struggle to: (1) understand a more sophisticated algorithm compared to what they had experienced in class; (2) find and fix bugs in the generated code; (3) learn about disciplinary concepts while they are also trying to fix the code; and (4) identify effective prompting strategies to instruct the ChatGPT how to complete the task. While LLMs show promise in supporting coding tasks for both professionals and students, using them requires strong background knowledge. When integrated into disciplinary courses, LLMs do not replace the need for effective pedagogical strategies. Our approach involved implementing a use-modify-create sequence, culminating in a productive failure activity where students engaged in conversations with the LLM encountered desirable difficulties. Our findings suggest that students faced challenges in trying to get a correct working code for FEA, and felt like they were teaching the model, which in some cases, led to some frustration. Thus, future research should explore additional forms of support and guidance to address these issues.

Traditional methods for teaching Internet-of-things (IoT) in engineering education often lack interactivity and hands-on engagement, limiting skill development. This study explores Augmented Reality (AR) based learning environment as a solution, enabling students to visualize real-time data flows, interact with virtual components, and configure IoT systems in a risk-free setting. Using tools like Unity 3D, Blender, Vuforia SDK, and Arduino IDE, an AR-based framework is developed and tested against traditional methods. The results show increased student engagement, knowledge retention, and skill acquisition, with a System Usability Score of 82.00%. The paper presents the framework's design, usability assessment, and comparative evaluation, highlighting AR's potential to enhance IoT education. It is observed that the proposed AR-based framework improves the skill retention by 36% over the traditional method. Further, the performance comparison of proposed method with traditional method is evaluated in terms of students' engagement, learning speed, and user satisfaction. In the end, the limitations of proposed study are addressed, and the future directions are presented.

The integration of artificial intelligence (AI) in higher education represents a transformative shift in the way teaching and learning are approached, offering unprecedented opportunities to enhance educational outcomes. One significant issue is the potential for bias in AI algorithms, which can perpetuate existing inequalities if not carefully managed. The objective of this study is to explore and evaluate the integration of AI in higher education to enhance teaching and learning processes. The study aims to identify the most effective AI tools and strategies for improving educational outcomes, assess their impact on student engagement and achievement, and provide actionable recommendations for educators and institutions. To effectively assess the integration of AI in higher education, a multifaceted data collection approach is essential. To ensure the successful integration of AI tools in higher education, a structured implementation plan is crucial. Enhancing teaching and learning involves a comprehensive approach that includes meticulous data collection, rigorous data analysis, strategic implementation and continuous improvement. The implementation phase requires thoughtful planning and execution, with a focus on refining AI systems based on feedback and performance metrics to ensure they effectively support educational goals. The findings show that AI integration in education has improved average grades to 88%, increased retention rates to 85%, and achieved 92% in content customisation and implementation using Python software. The future scope for integrating AI in higher education includes developing advanced AI tools that offer personalized and adaptive learning experiences, enhancing predictive analytics for student performance and retention, and fostering innovative pedagogical approaches through AI-driven insights.

This study investigates the potential of Multimodal Large Language Models to evaluate the quality of Unified Modelling Language (UML) class diagrams, with a focus on their ability to assess class structures and attribute information in alignment with object-oriented design principles. Thirty-four engineering students completed a design task involving the application of five object-oriented design principles known collectively as the S.O.L.I.D. principles (Single Responsibility, Open/Closed, Liskov Substitution, Interface Segregation, and Dependency Inversion). Their solutions were independently assessed by three expert instructors and four Multimodal Large Language Models: ChatGPTChatGPT-4, Gemini, Amazon AI, and Claude 3.5 Sonnet. Quantitative analysis compared AI-generated scores to instructor consensus ratings using inter-rater reliability metrics, while a grounded theory approach was used to qualitatively identify and classify AI evaluation errors. Results indicate that while MLLMs demonstrate promising partial scoring alignment with experts, they consistently exhibit significant limitations in semantic interpretation and evaluative reasoning, often leading to inconsistencies. These findings highlight that despite their potential, MLLMs are not yet reliable replacements for human expertise and underscore the critical need for improved model alignment with domain-specific assessment practices. They also suggest future directions for carefully integrated hybrid instructor-AI evaluation workflows in educational settings.

Higher education in cybersecurity faces significant challenges in developing practical and innovative offensive-defensive teaching cases. We present an automated framework for generating cybersecurity teaching cases using Large Language Models (LLMs), designed specifically for university-level cybersecurity education. The framework leverages the deep learning capabilities of LLMs and Artificial Intelligence Generated Content (AIGC) technology to enable intelligent construction and assessment of teaching cases. Our system allows instructors to automatically generate multidimensional teaching cases encompassing both known and potentially unknown security threats, based on parameters including network architecture, service configuration, security requirements, and network topology. Through prompt engineering techniques, the system enables fine-tuning of generated cases to accommodate diverse educational objectives and student proficiency levels. The framework incorporates an assessment module employing semantic analysis to provide automated multidimensional evaluation of student solutions, establishing a comprehensive pedagogical cycle. Empirical studies demonstrate that this framework significantly enhances the efficiency and quality of practical cybersecurity education, provides a replicable paradigm for vertical AI applications in higher education, and offers a novel approach to addressing resource constraints in university-level cybersecurity talent development.

With the rapid advancement of artificial intelligence, generative AI (AIGC) has emerged as a transformative tool in education, particularly in engineering disciplines where it demonstrates significant pedagogical potential. The CDIO (Conceive–Design–Implement–Operate) teaching model, rooted in experiential and project-based learning, emphasizes the development of students' integrated engineering competencies. However, engineering education remains challenging for many Chinese university students, despite the availability of online and collaborative learning resources, thereby underscoring the need for enhanced instructional strategies supported by advanced technologies. However, empirical research on the application of generative artificial intelligence in engineering education, particularly regarding its effects on individual psychological constructs, remains limited. This study, conducted over one semester in a data mining course, examines the impact of the AIGC-supported CDIO teaching model on students' computational thinking, learning motivation, engagement, and cognitive load. The participants included 76 s-year undergraduates from a teacher training university in China. The experimental group (n = 27) adopted the AIGC-CDIO teaching model, while Control Group 1 (n = 24) followed the traditional CDIO model, and Control Group 2 (n = 25) engaged solely in collaborative learning. Results from ANOVA analysis revealed that the experimental group demonstrated significant improvements in intrinsic motivation, behavioral and emotional engagement, and computational thinking abilities (including algorithmic thinking, critical thinking, and problem-solving skills), outperforming both control groups. Moreover, the experimental group exhibited significantly lower cognitive load. These major findings highlight the pedagogical effectiveness of the AIGC-CDIO approach in enhancing student engagement and reducing mental effort. The findings provide robust empirical support for the integration of AIGC into CDIO-based engineering education. This study contributes to the emerging literature on AI-assisted pedagogy by offering evidence-based insights into the interplay between technological mediation, psychological factors, and cognitive skill development. Implications for future research include deeper investigations into the mechanisms linking AIGC use to learning outcomes, longitudinal tracking of computational thinking development, and the refinement of adaptive instructional models across diverse learner profiles.

This paper tackles the common challenges in traditional manufacturing education, such as high equipment costs, significant operational risks, and limited hands-on opportunities. It proposes the design and development of a virtual simulation cognition system for a cylinder head production line, built on Unity3D. The system fully leverages virtual simulation technology to accurately replicate the complete cylinder head manufacturing process, encompassing critical stages like casting, machining, and assembly. In its design, the system deeply integrates situated cognition learning theory and educational game theory. By creating realistic operational scenarios, incorporating motivational mechanisms, and providing immediate feedback, it establishes a highly immersive and interactive learning environment. This empowers students to engage in self-directed learning and practical operations within a safe, controlled virtual space, thereby fostering the internalization of knowledge and the mastery of skills. The paper meticulously details the system's development process, core functional design, instructional content arrangement, and an analysis of teaching feedback. Our aim is to use technological means to optimize traditional manufacturing education, enhance teaching efficiency, and improve students' practical abilities, ultimately offering an innovative digital solution for nurturing talent in the manufacturing industry. This study begins by analyzing the application background of virtual simulation technology in intelligent manufacturing and educational training. It clarifies both the academic significance and practical necessity of designing and implementing a virtual simulation-based teaching system. The system's requirements analysis thoroughly considers theories like educational game theory and situated cognition learning. Based on actual teaching needs, it delineates functional modules including equipment cognition, layout construction, and task assessment. Compared to existing similar systems, this study optimizes the realism of the simulation, the depth of interaction, and the integration of learning theories, striving to provide a more effective learning experience. Throughout the development process, the system underwent multiple rounds of testing and verification, ensuring its functional completeness and performance stability. Ultimately, the virtual simulation system successfully achieved dynamic simulation of an automotive engine cylinder head production line, providing students with a realistic and highly interactive educational experience that helps them better understand and master complex manufacturing operations.