Korea University of Technology and Education expands AI-powered classrooms to sharpen teaching and learning

Korea University of Technology and Education is expanding its network of "AI learning analysis rooms," which use artificial intelligence to analyze instructors' speech and student engagement in real time. The university plans to grow the number of such rooms from three to 10 by year-end, and will apply the education model — once validated at the undergraduate level — to lifelong vocational training programs for working professionals.

The AI learning analysis rooms, housed in the Dadam Future Learning Center on campus, collect and analyze data on instructor behavior and student participation during class using computer vision, speech-to-text recognition and generative AI technology.

Six cameras mounted on classroom ceilings capture behavioral and audio data from both instructors and students. Instructors can monitor their own speech content, volume, pace and key learning keywords in real time on a large display at the back of the classroom. The system also provides engagement data derived from analysis of student movement and gaze patterns.

Instructors can use the analysis results to immediately adjust their teaching approach and the difficulty level of the material. When AI generates quiz questions based on lesson content, instructors can gauge student comprehension through the responses. If the error rate on a particular question is high, the instructor can revisit the concept students are struggling with.

After each class, the AI extracts key learning keywords and provides related questions and explanations. The system also generates three types of analysis reports — one for administrators, one for instructors to reflect on their teaching, and one for students reviewing material. Students can work with an AI tutor to tackle problems suited to their level of understanding or revisit content they have not fully grasped.

The university piloted the AI learning analysis rooms in the second semester of 2024 across four courses, including Introduction to Big Data and Data Visualization. The program has since expanded to more than 10 courses spanning engineering, humanities and general education. The university is also accumulating learning-pattern data across different class formats, including lectures, lab sessions and interdisciplinary discussions.

Byun Hae-won, a professor in the Future Convergence School who has taught in the AI learning analysis rooms since 2025, said the system has made individualized instruction possible through real-time data analysis and AI-driven two-way interaction. "Student satisfaction has risen by about 20 percent compared with conventional classes, and achievement and activity data can now be fed directly into the design of the next lesson," Byun said.

Data generated in the classrooms is linked to K-LXP, an AI-based student learning and growth support platform the university launched in March. Through the platform, students receive personalized learning and career guidance tailored to their goals and skill levels. A university-wide integrated AI chatbot gives students access to generative AI tools including ChatGPT, Gemini and Claude.

The university plans to add seven more AI learning analysis rooms this year, bringing the total to 10. After using undergraduate education as a testbed to verify effectiveness, it will extend the model to online education, on-the-job training for working professionals, and instructor training for vocational programs under the broader lifelong vocational development umbrella.

The investment in education is also translating into employment outcomes. In graduate employment statistics released by the Ministry of Education late last year, the university posted an employment rate of 82.8 percent — the highest among four-year universities nationwide with 500 or more graduates. That figure is 20.0 percentage points above the national average of 62.8 percent.

The share of graduates who remained employed for 11 months after joining a company stood at 89.6 percent, also 10.0 percentage points above the national university average. Annual education spending per student is 46.08 million won ($33,400), about 2.5 times the national university average of 18.33 million won.

University President Yoo Gil-sang said the AI learning analysis rooms serve as "a data compass for instructors to improve class quality and an educational infrastructure that supports personalized growth for students." He added that the university would take a leading role in developing AI talent for the nation, encompassing both undergraduates and working professionals.