Jeonbuk National University (JBNU) Department of Software Engineering students recently demonstrated the department's strong research competitiveness by winning two Gold, two Silver, and four Bronze awards at the summer conference and student paper competition of the Korean Institute of Information Technology held in Jeju.

The conference featured a wide range of research focused on cutting‑edge fields such as artificial intelligence (AI), robotics, Edge AI, and multi‑agent systems. Participating students were recognized for their creative ideas and practical implementation skills.

The Gold awards went to the team “Multi‑Agent Based Expression Habit Analysis System” (Park Chanyu, Kim Junwon, Im Changwoo, Hwang Junwon) and the team “Modumoa STF·RAG Role‑Separation‑Based Conversational Product Recommendation System” (On Geon, Lee Jehun, So Sangyun, Harfi).

The awarded student team proposed a technology that analyzes voice and nonverbal behavior in real time during interviews and presentations to detect repetitive expressions and suggest improvements. Their use of a multi‑agent architecture was credited with enhancing contextual understanding. The other Gold‑winning team gained attention for implementing a conversational system that applies separate STF and RAG structures to more precisely interpret user queries and improve recommendation accuracy and response quality.

Silver awards went to “LLM‑Based Asset Management Shell Generation System for Digital Twin Resource Registration” (Hwang Haegyu, Woo Hyeri, Kim Minjae, Hwang Juwon) and “Multi‑Agent System Considering Clinical Contexts in Healthcare” (Kwon Yewon, Kim Hyunji, Kim Jaehong, Lee Junhyung). These studies were evaluated as increasing applicability in industrial and medical fields by respectively presenting an automated asset information structuring technique using LLMs and a collaborative AI architecture that reflects the characteristics of medical environments.

In addition, four teams won Bronze: a hybrid reinforcement learning method for an Eye‑In‑Hand Robot System for non‑adhesive part loading (Yang Jaewoo, Yoo Seungbin, Jung Jaewon, Lee Jueun, Suhbat); an Edge AI‑based early warning system for infant falls at home (Lee Jaehun, Kim Seonggwang, Shin Junhyun, Kang Taehun, Kim Daejin); an LLM‑based execution control system to enhance the reliability of the ARGUS AI Agent (Lee Sangeun, Hwang Soyun, Kim Jisoo, Kim Yunsu); and a UI Sketch‑based dynamic software prototype generation system (Koo Hyunjoo, Kim Hyowon, Woo Heeseok, Kim Donggui).

Notably, the combination of robot vision and reinforcement learning for industrial automation, real‑time fall detection systems that consider personal data protection, control techniques that enhance AI agent stability, and UI‑sketch‑based automatic prototype generation were highlighted as practical research with high potential for field application.

The winning students said, 'Through this competition we accumulated research experience in the latest AI and software technologies and were able to further strengthen our team‑based problem‑solving skills and practical capabilities.'

Professor Kim Sun‑tae (Department of Software Engineering), who supervised them, said, 'It is highly meaningful that the students demonstrated creative problem‑solving abilities based on the latest technologies,' and added, 'We will continue to cultivate competitive talent through practice‑oriented research linked with industry.'