Since the implementation of AI digital textbooks in 2025, digital infrastructure has been expanded, but the quality of educational content has not kept pace. In particular, there is a lack of interactive and highly immersive content that leverages the strengths of the digital environment.

01 | Online Survey and Teachers Interview

What do you think is the most important reason why AI digital textbooks are not as effective as expected? (Multiple answers possible) (Multiple responses allowed)

Digital Devices ≠ Digital Learning

Digital devices are prepared, but the quality of educational content is insufficient

I conducted a comparative analysis of student and teacher experiences in traditional textbook classes versus digital textbook classes. While digital textbooks provide interactive experiences, they actually reduce students' active inquiry and increase teachers' lesson preparation burden.

Currently used digital textbooks are reducing discussion experiences with peers and replacing them with quiz content.

Traditional Textbooks

Digital Textbooks

I conducted a comparative analysis of student and teacher experiences in traditional textbook classes versus digital textbook classes. While digital textbooks provide interactive experiences, they actually reduce students' active inquiry and increase teachers' lesson preparation burden.

Digital textbooks provide students with interactive experiences, but they also increase the burden of device management for teachers.

Traditional Textbooks

Digital Textbooks

As digital textbooks become more common in science classrooms, efficiency gains often come with tradeoffs—reducing peer interaction, limiting inquiry time, and shifting learning toward answer-checking rather than exploration.

01 | Problem Definition

02 | Research Question

"How can we design digital textbook content to reduce the burden on teachers and students and transform passive learning into active inquiry experiences?"

After identifying the key constraints in the classroom setting and the gaps in student learning, I created an MVP to rank and prioritize features based on their importance. The main focus was on features that directly support inquiry-based science learning while still being practical for everyday classroom use.

I organized the problems that teachers and students face into three design directions aligned with the MVP list. At the same time, I divided 14 experiments into four experiment types and created reusable XR templates for each prototype.

Reflecting on insights from earlier research and the defined design directions, I mapped an ideal teacher journey to show how lesson preparation, delivery, and assessment change with the introduction of digital textbooks.

Teacher Flow

Teachers shift from troubleshooting devices to facilitating inquiry by using step-by-step guides and built-in measurement tools.

Building on the proposal from same design directions, I mapped an ideal student journey to illustrate how learning activities, interaction, and feedback differ between traditional classrooms and digital textbook-based learning.

Student Flow

Students move from reading and clicking to manipulating parameters, observing change, and testing hypotheses.

To implement experiments from paper-based textbooks as web-based interactive simulations, I modeled and simulated how the experiments work and rendered them in real-time using 3D/graphics technology.

Sketching

3D Modeling

Computer Graphics

I created video data by filming scenes of grilling meat and fish with a thermal camera.

Based on the video data, I developed a thermal camera mode simulation using GLSL shaders.

Stakeholder feedback from MiraeN’s editorial and product teams informed targeted UI refinements, with a focus on readability, clarity, and usability for elementary students using tablets in classrooms.

Based on earlier analysis of classroom workflows and inquiry breakdowns in digital textbooks, I designed interaction patterns around recurring inquiry structures found across science units, including time-based change, state comparison, and spatial observation.

Group 1 | Time Change Simulations

Slider → Scene/Graph changes → Read Values

Learners adjust time by tapping or dragging the slider horizontally or vertically, use play and pause to fine-tune moments, zoom to read angular values, and record sun altitude, shadow length, and temperature at fixed intervals.

Group 2 | State Change Simulations

Tap 3D Objects → State Changes → Read Outcomes

Learners tap 3D components to set up comparable experimental groups, keep variables constant, and run the same experiment across conditions to compare outcomes and reason about controlled variables.

We implemented interaction patterns that operationalize inquiry by allowing learners to change conditions and observe outcomes.

Group 3 | Comparison & Classification

Drag Objects → Observe Differences → Review Results

Learners create comparison groups, move objects between conditions, observe behavioral differences, and review results after each observation cycle.

Group 4 | 3D Structure & Observation

Rotate & Zoom → Reveal Hotspots → Read Explanations

Learners rotate and zoom the 3D scene to inspect structures from multiple angles, toggle hidden elements, and access contextual explanations embedded within the model.

I grouped the curriculum into four experiment types and created interaction and UI design guides.

XR Science Museum begins each activity by orienting learners to the experiment, clarifying its goal and overall flow before hands-on exploration starts.

I created an activity guide for each experiment. This activity guide can access by 'Activity Guide' button on the intro screen.

During curriculum-based experiments, the system visualizes changes in real time and provides continuous feedback that helps students track progress, interpret results, and recognize when a trial is complete.

Time-based simulation supports inquiry by allowing students to observe how variables change over time rather than relying on static snapshots. By controlling time directly, students can pause, rewind, and compare moments to reason about change through repeated observation and measurement.

State-change type of simulations let students directly manipulate components to compare discrete conditions side by side. By toggling elements such as switches, batteries, or bulbs, students can observe immediate cause-and-effect relationships between system states.

Comparison and classification simulations type of simulation allows students directly manipulate objects to test how they behave under different conditions. By dragging items between containers, removing them, and trying again, students can compare outcomes, identify patterns, and build their own sequences of classification through repeated testing.

3D structure and observation simulations allow students to explore spatial relationships by viewing phenomena from multiple perspectives. By rotating the scene and adjusting light sources, students can observe how visibility and alignment change across positions, helping them reason about spatial structure and relative motion.

By analyzing recurring patterns across curriculum-based experiments, I translated diverse science content into four reusable XR interaction templates—creating a scalable system that supports inquiry, comparison, and observation across grade levels.

Each category informed a consistent XR layout and simulation logic.

XR Science Museum will launch in March 2026 and be deployed across multiple classrooms. This rollout enables validation beyond prototyping—allowing us to evaluate learning impact, classroom usability, and system reliability in real instructional contexts. Post-launch, we will analyze usage data, teacher feedback, and student work to assess the following questions.

01 | Does XR interaction support conceptual understanding—not just engagement?

- Whether students can describe, predict, and compare phenomena using their own words based on observed outcomes
- Whether XR interactions help reduce common misconceptions found in textbook-based instruction

02 | Does the system reduce instructional friction for teachers?

- How much the system reduces time spent on setup, explanation, transitions, and resets compared to existing materials
- Which workflows or features become classroom bottlenecks and should be prioritized for iteration

03 | Is the system reliable under real classroom constraints?

- Whether loading time, frame rate, and interaction responsiveness remain stable across devices and networks
- How performance issues affect learning behaviors such as drop-off, retries, and task abandonment

2026 MiraeN Science Textbook Overview Video

Together, these evaluations will inform how XR can move from experimental novelty to a dependable learning tool in everyday classrooms.