A Retrieval-Augmented Generation Platform for Construction Education: Development and User Study

Document Type

Conference Proceeding

Publication Date

1-1-2026

Abstract

The continued digitalization of construction practice, including the expanded use of BIM, automation, and data-centric workflows, has increased the demands placed on construction management education. As curricula adapt to these developments, both instructors and students face growing challenges in organizing and interpreting course content during independent study. Traditional slide-based review materials provide limited support for individualized learning and offer few mechanisms to help students reconstruct conceptual relationships. Although general-purpose generative AI tools can provide on-demand explanations, they are not aligned with course-specific terminology and rarely offer verifiable links to instructional materials, which limits their suitability for guided review. Guided by constructivist perspectives on explanation and information integration, this study develops and evaluates a course-aligned review platform that uses Retrieval-Augmented Generation (RAG) to deliver contextualized explanations grounded in lecture slides. A formative user study was conducted in which participants viewed a common instructional video and completed a review session using either the AI-driven platform or conventional slides, followed by a user-experience questionnaire and a post-review assessment. The AI review condition received higher ratings on key UEQ dimensions, including Perspicuity, Efficiency, Dependability, and Attractiveness, and showed a modest advantage on explanatory performance. The findings indicate that domain-bounded, citation-aware generative AI can enhance transparency and support more interpretable review practices in construction education.

Publication Title

Proceedings of the International Symposium on Automation and Robotics in Construction

ISBN

[9780645832235]

Link to Full Text

Share

COinS