UST: Achievements of National R&D Utilizing AI Becoming Visible

The University of Science and Technology (UST) announced on March 11 that the achievements of national research and development (R&D) utilizing artificial intelligence (AI) are becoming increasingly visible.

UST operates the "AI Curriculum Completion Certification System." This system certifies non-AI major students who have completed AI-related courses, verifying their ability to generate research achievements using AI and evaluating their AI competencies.

UST building exterior. University of Science and Technology

Last year, a total of 26 students received certification, more than a fivefold increase compared to the previous year (5 students). Notably, during the second round of certification reviews last year, 18 students were certified, and among them, 9 received "excellence certification" for having strategically applied AI as a core tool in their research process and achieving results of global significance.

These students who received excellence certification strategically integrated AI technologies such as large language models (LLMs) based on text learning, including Gemini and ChatGPT; vision foundation models (VFMs) trained on vast amounts of image and video data; and knowledge graph-based retrieval-augmented generation (GraphRAG), which connects complex relationships between pieces of information to understand the bigger picture, into their specialized research fields.

UST explained that these students not only used AI as a simple data analysis tool but also designed the entire research problem-solving structure based on AI, leading to outstanding results.

For example, Hyunwoo Seo, a master's student majoring in Aerospace System Engineering at the Korea Aerospace Research Institute (KARI) School, utilized a vision foundation model to accurately detect ships as contextual information in satellite images and improved the ability to describe related situations precisely in text.

The research findings (paper) were presented at the international academic conference in the field of remote sensing and satellite imaging, "IGARSS 2025 (IEEE International Geoscience and Remote Sensing Symposium 2025)."

Wondesen Workneh Ejerssa, a doctoral student majoring in Energy-Environment Convergence at the Korea Institute of Science and Technology (KIST) School, achieved the feat of publishing as first author in two SCIE journals ranked in the top 9% by JCR, utilizing AI technology.

This student derived optimal application criteria reflecting the characteristics of each pollutant in an ultraviolet water treatment process that removes micro-pollutants from water using ultraviolet rays. The research outcomes were published consecutively in the chemical engineering journal "Separation and Purification Technology" and the environmental science journal "Chemosphere."

Hyunwoo Cha, a master's student majoring in Construction Environmental Engineering at the Korea Institute of Civil Engineering and Building Technology (KICT) School, applied knowledge graph-based retrieval-augmented generation (GraphRAG) technology to complex three-dimensional construction design data (Building Information Modeling). He developed an automated search model that enables AI to create appropriate queries and retrieve desired construction information even when users are unfamiliar with complex search terms.

Based on this, he designed a construction material recommendation system using agent-based reinforcement learning and presented his research as the first author at international academic conferences, including ICAT 2024.

Reflecting on these achievements, UST plans to make AI curriculum completion mandatory for all new students starting next year. The university also intends to expand required AI courses and AI MOOCs, as well as establish new AI research groups involving both major and non-major faculty members, so that students can lead AI-driven research in their respective fields.

UST President Daeim Kang stated, "AI is a next-generation core technology driving innovation across all academic disciplines and industries," adding, "We will continue to develop academic systems and provide support so that UST students can strategically design and utilize the latest AI technologies in their own specialized research fields to achieve world-class research outcomes."

This content was produced with the assistance of AI translation services.