AI-Based Real-Time Interactive VR Content for Future Teacher Training

AI-Based Real-Time Interactive VR Content for Future Teacher Training

OWLIX AI VR Content — Busan National University of Education 1st TSR Pilot

" Through AI-based interactive education services, we effectively assess and improve pre-service teachers’ instructional guidance, communication, and problem-solving abilities. "


AI evaluation based on the trainee’s spoken content	Questions from AI students, responses from trainees	Question types and evaluation criteria customizable by the instructor	Future-oriented classroom design

INTRO

Lack of practice opportunities · Difficulty in objective evaluation · Absence of instant feedback · Ability to handle diverse situations · Communication training

Traditional pre-service teacher education has been theory-centered, with limited opportunities to practice with actual students. It has been difficult to objectively evaluate communication skills, respond to the many variables that arise during real classes, and provide immediate feedback to improve instructional competence.


Content screen shown during actual learning	Scene from an external viewpoint

SOLUTION :

The 1st TSR Pilot addresses these issues by linking an Unreal Engine–based virtual environment interface with a central management server. Trainees interact with AI students connected to a local AI pipeline in a virtual classroom. Instructors create and control classes in real time via the web, and all data is aggregated on a central server (SQLite) to generate AI evaluation reports.

1. Central Management Server (Professor PC)	RESTful API–based. Class orchestration, PDF distribution, real-time monitoring, and final report integration
2. Integrated Client (Trainee PC)	Composed of a Python client and an Unreal Engine client; the Python client acts as a broker between the central server and Unreal Engine
3. Local AI Pipeline	Processes STT, LLM, and TTS on the client PC
4. Virtual Environment Interface	Unreal Engine–based user interface; interact via HMD and controllers
5. AI Evaluation System	Generates detailed reports including total score, strengths, and areas for improvement based on the pre-service teacher’s lesson and responses

1st TSR Pilot Components

Samwoo Immersion’s 1st TSR Pilot consists of a web-based central management server for evaluators (professors) and a VR integrated client for trainees (pre-service teachers).


01. VR Access & Login After the trainee puts on the HMD and runs the content, log in using the student ID issued by the instructor	02. Enter Classroom & Verify Settings Enter the virtual classroom and check the teaching materials (PDF) and the podium buttons	03. Class Execution (Recording) Press 'Start Recording' to record the lesson demonstration through the HMD microphone

04. AI Student Questions When the lesson content is analyzed, AI students raise their hands and generate questions	05. Answer & Evaluation Press 'Start Recording' and answer the AI students’ questions. The AI evaluates the responses	06. End Class & Generate Report When 'End Session' is pressed, the AI generates a final evaluation report based on the entire class and responses, then the content closes automatically

Samwoo Immersion 1st TSR Pilot

AI-based real-time interaction and automated evaluation

1. AI Pipeline Integration
A Python-based local AI pipeline connects Unreal Engine with the central server to generate questions and perform evaluation

2. Centralized Data Management
The instructor’s central management server integrates class creation, student management, real-time monitoring, and final report collection

3. Instructor Customization
Instructors upload PDF lesson plans and preset AI question types and evaluation criteria

Key Features

Bidirectional interaction with AI students, real-time instructor control, and AI auto-generated evaluation reports to strengthen pre-service teachers’ practical capabilities.

AI Student Question Generation	Instructor Web Dashboard	AI Auto Evaluation Report

AI analyzes lesson content to auto-generate questions and conduct interactive dialogue with trainees	Via the web: register trainees, create classes (PDF upload), manage sessions, and configure AI evaluation criteria	At session end, AI automatically generates a detailed report with total score, overall evaluation, strengths, areas for improvement, and interaction logs