Online Session 3
Order	Time	ID	Full information
1	13:00-13:15	IT1069	BilimALL AI: An Intelligent Platform for Integrating Computer Science and Natural Sciences in Digital Education Author(s): Meruert Yerekesheva,  Ryskul Zhakhina, Adilet Darmenov, Anar Tashimova, Moldir Muratova Presenter: Yerekesheva  Meruert, K. Zhubanov Aktobe Regional University, Kazakhstan Abstract: In the modern education system, the integration of Artificial Intelligence (AI) technologies into the learning process has become one of the key priorities. This study focuses on the development of the BilimALL AI platform, which was created based on the concept of connecting informatics with natural science subjects and enhancing digital learning practices. BilimALL AI is an intelligent system that identifies interdisciplinary links between informatics and natural sciences, assisting teachers in automatically generating educational materials and visual slides
2	13:15-13:30	IT2126	Deep Teaching Reform and Practice of Energy Artificial Intelligence Curriculum Based on Digital-Twin-Project-Driven Model Author(s): Xiaogang Liu, Yongfu Cheng, Yuanqi Gu, Xifeng Cao, Yuhang Wang, Chenxi Zhao Presenter: Xiaogang Liu, Harbin University of Science and Technology Abstract: To meet the urgent demand for AI talent in energy systems under the “dual carbon” strategy, this paper proposes a digital-twin-enabled, project-driven teaching model for a course on Energy AI. The model addresses a structural “triple disconnect” in current education by targeting key deficiencies, including the gap between theory and practice and the lack of systems thinking and engineering skills. It builds an integrated “data–model–decision verification” framework, supported by a three-step experimental system based on energy-system digital twins, and embeds a project-based learning (PBL) project centered on a virtual power plant (VPP) operation scenario. Within this project, students develop forecasting models for distributed energy resources, design market bidding and dispatch strategies, and validate their effectiveness in a grid-interactive digital-twin VPP environment. Comparative teaching experiments and multidimensional evaluation based on five core competencies demonstrate that the model significantly enhances students’ complex problem-solving ability, innovation awareness, and teamwork skills, and offers a replicable pathway for cultivating high-quality engineers for the new era of energy and power systems.
3	13:30-13:45	IT2154	Design and Practice of an Interdisciplinary and Enterprise-Integrated Training Model for Postgraduate Engineering Students Author(s): Ye Dai, Deyu Meng, He Hao, Long Li, Shuang Yu, Wenyin Qu Presenter: Ye Dai, Harbin University of Science and Technology, China Abstract: This paper proposes and implements a closed-loop interdisciplinary and enterprise-integrated training model for postgraduate engineering students. The model is designed to bridge the gap between academic learning and industrial practice by structurally integrating enterprise projects into the curriculum, supervision, and assessment processes. It consists of four coupled components: enterprise-aligned interdisciplinary electives, enterprise real-project-driven project-based learning (PBL), dual-supervisor collaborative governance, and a dynamic evaluation and feedback mechanism. A digital platform serves as the operational hub to connect students, courses, supervisors, and projects. The model was piloted across three engineering colleges at Harbin University of Science and Technology in collaboration with six industrial partners. Evaluation data from two student cohorts show significant improvements in students’ interdisciplinary competencies, enterprise acceptance of deliverables, and job offer rates, demonstrating the effectiveness of the approach in enhancing the employability and industrial readiness of engineering graduates.
4	13:45-14:00	IT3209	Research-Led Curriculum Innovation under Medical Engineering Convergence: Redesign and Practice of a Course in Clinical Medical Data Analytics Author(s): Wei Liu, Xiaoling Li, Jialun Lin, Quanhai Zhang, Yuanbo Yu Presenter: Wei Liu, Hainan Medical University, China Abstract: The rapid transition toward data-driven and intelligent healthcare has imposed new competency requirements on medical education, particularly in the integration of clinical reasoning with data analytics and engineering methodologies. Conventional teaching models in clinical data analysis often suffer from a separation between research and instruction, simplified datasets, and limited exposure to real-world clinical complexity. This study reports a research-led teaching reform of the course Clinical Medical Data Analytics, guided by the principle of deep medical–engineering integration. Instead of organizing instruction around isolated analytical techniques, the course was redesigned around authentic clinical research problems derived from ongoing projects on sepsis-related complication prediction. Real-world intensive care databases, such as MIMIC-IV, and advanced analytical approaches—including dynamic early warning modeling and model interpretability techniques—were systematically embedded into course modules and project-based learning tasks. A progressive instructional framework was developed, encompassing clinical problem formulation, medical data governance, intelligent modeling, and clinical value assessment. Teaching effectiveness was evaluated through learning outcomes, student feedback, teaching supervision, and research spillover effects. Over two academic years, the reformed course demonstrated significant improvements in student engagement, interdisciplinary competence, and research-oriented learning outcomes. The proposed model illustrates a sustainable “research–teaching symbiosis” paradigm and offers a transferable reference for interdisciplinary curriculum development in intelligent medical education.
5	14:00-14:15	IT3277	A Dual-Driven Model for Cultivating Innovative Control Postgraduates: Integrating Interdisciplinary Knowledge with Generative AI Author(s): Junjie Liu, Lipeng Zhang, Yuehui Ji, Qiang Gao Presenter: Lipeng Zhang, Tianjin University of Technology, China Abstract: The rapid evolution of the intelligent era presents significant challenges to postgraduate education in control disciplines, notably rigid disciplinary boundaries and a lag in incorporating cutting-edge technologies. To address these issues, this paper proposes a novel training model designed to enhance the innovation capability of postgraduate students. This model is propelled by a dual-drive mechanism: "Interdisciplinary Integration" and "Generative Artificial Intelligence." It achieves its objectives through a two-pronged approach: first, by deeply embedding generative AI tools into the core curriculum to broaden students' knowledge base and reinforce systems thinking; and second, by systematically integrating these tools into the entire scientific research training workflow to improve research efficiency and deepen innovative outcomes. The ultimate goal is to cultivate a new generation of control specialists who are proficient in leveraging intelligent technologies to solve complex, interdisciplinary problems at the forefront of the field. This work aims to provide a replicable and scalable paradigm for reforming postgraduate education in control and related discipline.
6	14:15-14:30	IT3314	The 'P-M-C-S' Closed-Loop Approach to Engineering-Problem-Driven Computational Methods Education in New Engineering Author(s): Xiaoying Zheng, Yuanyuan Li Presenter: Xiaoying Zheng, Wuhan Institute of Technology, China Abstract: Driven by the national New Engineering strategy and the integration of digital-intelligent technologies, engineering education is evolving, placing new emphasis on courses like Computational Methods. The latest Engineering Education Accreditation Standard (2024 Edition) explicitly elevates computation to a fundamental pillar, alongside mathematics and science, requiring graduates to apply it to complex problems. To move beyond traditional, passive instruction, this work designs and applies a 'P-M-C-S' closed-loop teaching model. This problem-driven framework leverages digital tools to guide students through a complete cycle from Problem identification and Mathematical modeling to Computation and Simulation. Using the 'robotic arm inverse kinematics' problem as an exemplar, the paper details the course redesign and pedagogical process. Practice demonstrates that the model successfully motivates students, strengthens their grasp of algorithmic principles in practical contexts, and systematically cultivates the multifaceted computational competency mandated by modern accreditation and industry needs.
7	14:30-14:45	IT3327	AI-enhanced Design of Interdisciplinary Thematic Learning in High School Biology: From the Perspective of Integrated STEM Education Author(s): Chen Rutian, Feng Chunyan, Shen Xinyue, Zhou Zhe Presenter: Chen Rutian, The University of Hong Kong, Hong Kong Abstract: In the context of Industry 4.0, traditional single-subject teaching is unable to meet the demand for cultivating innovative and interdisciplinary talents. Integrated STEM education, by placing disciplinary knowledge in real problem-solving contexts, provides good support for the cultivation of 21st-century competencies. This study adopts a mixed-methods research design to verify the effectiveness of the AI-enhanced design pathway for interdisciplinary thematic learning in high school biology and its teaching practice in improving students’ disciplinary core competencies and interdisciplinary competencies. The results indicate that the experimental group (EG) achieved significant improvements in disciplinary core competencies, including scientific concepts, scientific thinking, scientific inquiry, and social responsibility, as well as in interdisciplinary competencies such as innovation and collaboration. Furthermore, the proposed design pathway effectively enhances student initiative and engagement, addressing their diverse inquiry needs. This study offers new insights for implementing interdisciplinary thematic learning in high school biology and provides practical evidence for future research in related fields.
8	14:45-15:00	IT4446	Exploring an AI-Empowered “1+M+N” Advanced Mathematics Teaching Model Based on a Three-Tier Collaborative Architecture Author(s): Lina Jia Presenter: Lina Jia, Zaozhuang University, China Abstract: Addressing key challenges in Advanced Mathematics education—its abstractness, insufficient student interest, and lack of humanistic cultivation—this study constructs an innovative AI-empowered “1+M+N” teaching model. Here, “1” is the main thread of mathematical history, “M” denotes core mathematicians’ stories, and “N” represents fundamental mathematical ideas. Based on a three-tiered architecture (resource, technology, teaching layers), a multimodal resource repository and AI-driven functional modules are built to achieve in-depth integration of AI and teaching practice. A case study on “The Concept and Applications of Derivatives” among engineering undergraduates verifies the model. The results show that it effectively resolves abstract teaching difficulties, improves student engagement and knowledge mastery, and facilitates the cultivation of mathematical literacy and inquiry abilities. Moreover, the three-tiered architecture’s synergy ensures sustainable model implementation, providing a feasible path and practical reference for advanced mathematics education reform in the AI era.
9	15:00-15:15	IT4449	Bridging the PBL Gap: A Cloud–Edge–Device Conversational AI Platform for STEM Inquiry and Critical Thinking Development Author(s): He Huang, Qianyi Wu, Zehui Zhan Presenter: He Huang, Shenzhen Luohu Foreign Languages Junior High School, China Abstract: Project-based learning is vital for developing students’ core disciplinary and interdisciplinary skills. Traditional PBL models face challenges such as limited resources, delayed formative assessment, and lack of personalized support. To address these, we developed a Conversational AI-Enabled STEM Platform based on a Cloud–Edge–Device architecture. The cloud layer enables regional resource sharing; the edge layer provides low-latency LLM services and real-time AI inference via local GPUs and RK3568 devices, ensuring privacy; and the device layer integrates tablets and IoT sensors for immersive labs. Core features include AIGC-based content creation, intelligent micro-lectures, multi-source AI assessment, and an incentive scoring system. An interdisciplinary pilot found that students’ technological self-efficacy improved, especially regarding AI’s usefulness. Epistemic Network Analysis was applied to analyze academic achievement data and student-AI dialogue data. The results revealed that both high- and low-achieving students exhibited insufficient openness in their critical thinking dispositions, mainly due to knowledge gaps and an outcome-driven evaluation system. Compared to low achievers, High achievers were better at applying critical thinking in practice. The findings call for an iterative, concept-driven pedagogy, which using sub-problem scenarios, student-AI interaction modeling, and structured questioning, to foster critical thinking, especially in AI-supported learning environments, which offers key insights for improving the STEM platform, especially in supporting open inquiry, adaptive feedback, and guided reasoning.