Date Approved

5-29-2026

Graduate Degree Type

Thesis

Degree Name

Artificial Intelligence (M.S.)

Degree Program

College of Computing

First Advisor

Jiaxin Du

Second Advisor

Rahat Rafiq

Academic Year

2025/2026

Abstract

This thesis presents EduAtlas, a human-centered AI system that integrates a mobile lesson planning application with wearable smart glasses to support teachers in multilingual classroom environments. The system enables teachers to generate structured lesson plans with multilingual scaffolds and receive real-time instructional support during classroom sessions through a wearable interface. Unlike autonomous systems, EduAtlas is designed to operate collaboratively with teachers, who retain control over all AI-generated suggestions.

The lesson planning component uses large language models and multimodal generation to produce standards-aligned lesson plans, visual aids, and multilingual supports. The wearable component utilizes teacher-triggered audio capture to analyze classroom interactions and deliver concise, actionable guidance in real time. The system is designed with privacy-by-design principles, including teacher-triggered capture, no persistent audio storage, and no student identification.

This research is guided by Sociotechnical Systems Theory, Cognitive Load Theory, and Dual Coding Theory, which together inform a broader AI-Aided MLL Instructional Framework (AMIF). The thesis addresses five research questions related to theoretical grounding, teacher needs, system design, wearable feasibility, and deployment constraints.

The system was evaluated through a mixed-method design that included a co-design workshop with 55 educators, faculty-based evaluation of generated lesson plans, a controlled wearable latency pilot, and an expert pedagogical interview. Results indicate strong perceived usefulness for the lesson planner and demonstrate that the wearable pipeline can operate within classroom-relevant latency constraints under controlled conditions.

This work contributes a novel mobile--wearable AI architecture for classroom support and highlights the importance of teacher-in-the-loop design in educational AI systems.

Download

Share

COinS