KAIST College of Engineering

Development of a Radioisotope Battery Using Cyclotron Radiation of Beta Particles in a Magnetic Field

A novel radioisotope battery has been developed that converts radioactive decay energy into electricity by utilizing cyclotron radiation emitted by electrons in strong magnetic fields, demonstrating high conversion efficiency and long-term operational stability for diverse applications.

https://breakthroughs.kaist.ac.kr/sub02/view/id/11189

Nanoscale Property Maps for Resistive Switching: Co-localized Ionic–Electronic Dynamics

KAIST researchers used multi-modal atomic force microscopy to visualize co-localized changes in conductance, oxygen-vacancy concentration, and surface potential in titanium dioxide thin films, revealing the key resistive-memory mechanisms governed by coupled ionic and electronic processes

https://breakthroughs.kaist.ac.kr/sub02/view/id/11188

Dual-Action Hydrogel Offers New Hope for Rheumatoid Arthritis Treatment

A KAIST research team has developed a dual-functional hydrogel for rheumatoid arthritis that combines joint lubrication with immune regulation. The hydrogel is injectable, functions as a bio-lubricant, and alleviates immune cell–driven joint inflammation, thereby targeting key pathological features of rheumatoid arthritis.

https://breakthroughs.kaist.ac.kr/Korean/sub02/view/id/11187

Automating Steel Bridge Paint Inspection with Infrared and Vision

Professor Hoon Sohn’s research group has developed an automated inspection system that combines infrared and visual sensing to objectively evaluate steel bridge paint. The system measures coating thickness, detects damage, and assigns condition ratings, improving inspection safety, consistency, and maintenance decision-making.

https://breakthroughs.kaist.ac.kr/sub02/view/id/11186

Team KAIST takes third place at the World’s Top Autonomous Drone Racing Challenge in Abu Dhabi

Four graduate students of Prof. Shim, the “Godfather of Autonomous Drone Racing,” clinched third place, winning a prize of $112,250 at the world’s premier drone racing competition—where drones must navigate indoor gate courses using only an onboard monocular camera.

https://breakthroughs.kaist.ac.kr/sub02/view/page/2/id/6818

Method that reduces real-world AI training costs and its application to control infectious diseases

To address prevalent data quality issues in real-world AI training, this research focuses on three major challenges. First, it automatically detects and corrects label errors during training, reducing the need for manual data preprocessing. The related paper has had a significant academic impact, being cited over 1,200 times in the past two years. Second, it automatically infers missing labels in time-series data, significantly lowering the cost of manual label acquisition. Third, it removes redundant data and selects a core set of informative samples, achieving comparable model performance while reducing the training time by up to 90%. These technologies have been successfully applied to real-world social problems such as infectious disease prediction and economic impact forecasting. They have been granted patents in both South Korea and the United States.

https://breakthroughs.kaist.ac.kr/sub02/view/page/2/id/6817

GFlux: Trillion-edge scale graph processing with SSD-GPU acceleration

Processing large-scale graphs is increasingly vital in fields such as business and AI (e.g., vector database indexing). GFlux enables trillion-edge-scale graph processing on a single machine using SSDs and GPUs, supporting complex analytics such as triangle counting.

https://breakthroughs.kaist.ac.kr/sub02/view/id/6816

Wearable hyperspectral photoplethysmography for the continuous monitoring of exercise-induced hypertension

A wearable hyperspectral PPG module using a compact and high-resolution microspectrometer allows continuous, cuff-free blood pressure monitoring with high accuracy by capturing spectral PPG signals and physiological changes during exercise.

Wearable hyperspectral photoplethysmography (HS-PPG) module for blood pressure monitoring. The HS-PPG module integrates a high-resolution microspectrometer and allows the acquisition of the hyperspectral range of PPG signals. Blood pressure levels estimated using the HS-PPG exhibit high accuracy, with R2 values up to 0.95.

https://breakthroughs.kaist.ac.kr/sub02/view/id/6815

Visible Light : A tool for fusion energy development

A new visible Bremsstrahlung system was developed for the VEST Tokamak to monitor impurity content in its plasmas. This system will contribute to a better understanding of impurity behavior and its impacts on fusion plasma performance.

Developed visible Bremsstrahlung system in the VEST Tokamak and its measurements results.

https://breakthroughs.kaist.ac.kr/sub02/view/id/6814

From local to route-wide V2X: Smarter driving with predictive control

A new approach from KAIST enables electric vehicles to make smarter driving decisions by looking far ahead on the road. Using route-wide information shared through V2X, the system helps vehicles save energy—almost as efficiently as if they could see the future.

V2X communication helps electric vehicles look ahead—enabling smarter, more energy-efficient driving.

https://breakthroughs.kaist.ac.kr/sub02/view/id/6813