生成AIを活用したプロジェクトベースドラーニング(PjBL)の可能性:「社会実装・地域創生人材副プログラム」に向けた授業設計の提案
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
This paper examines how generative AI can be integrated into Project-Based Learning (PjBL) in engineering and information science universities in ways that enhance students’ problem-finding abilities without undermining learner agency. In preparation for the planned “Social Implementation and Regional Revitalization Human Resource Subprogram” at Kyushu Institute of Technology, the study explores the potential of an instructional design that integrates problem discovery grounded in the humanities and social sciences with the development of engineering-based solution concepts. In conventional engineering-oriented PjBL, problem solving has tended to be foregrounded, while problem finding has received relatively little attention. However, PjBL aimed at regional revitalization and social implementation requires problem setting that considers social contexts such as institutions, culture, history, and stakeholder relationships. At the same time, it is difficult to provide novice students with sufficient experience in humanities- and social-science-based methodologies within the limited time available in regular courses. To address this constraint, this study focuses on the use of generative AI as a potential means of support. Drawing on prior research, a theoretical framework is developed to prevent excessive reliance on AI and the substitution of students’ own thinking. The study concludes that generative AI should be positioned not as an agent that performs problem finding on behalf of students, but as a “companion” that supports the examination and expansion of problems already defined by learners. Specifically, two instructional design principles are proposed: (1) establishing a phase in which students explicitly articulate their own problem definitions before using AI, and (2) requiring students to document the purposes of AI use, their prompts, and the reasons for adopting or rejecting AI outputs, thereby encouraging metacognitive reflection on the problem-finding process. Based on this framework, a small-scale PjBL was implemented, and eight submitted AI usage logs were analyzed. The results indicate that students primarily used generative AI for peripheral tasks such as supporting idea generation, collecting general information, and drafting or revising text; no instances were observed in which AI was entrusted with the core of problem setting. However, AI used for revising, deepening, or restructuring problem frames was limited. These findings suggest that, for generative AI to function more effectively as a “companion” in problem finding, instructional activities must be designed so that such use becomes a necessary and meaningful part of the learning process.
Fields of Study & Subjects
Problem and Project Based Learning
Artificial Intelligence in Education
Design Education and Practice
Publication details
- Journal / source
- Institutional Repositories DataBase (IRDB)
- Volume
- 10
- Pages
- 37–50
- Publication date
- 31 March 2026
- Language
- JA