Sungkyunkwan University Undergrad Researcher Develops Drone AI Identification Semiconductor Tech… Published in World-Renowned Academic Journal
en-GBde-DEes-ESfr-FR

Sungkyunkwan University Undergrad Researcher Develops Drone AI Identification Semiconductor Tech… Published in World-Renowned Academic Journal


Sungkyunkwan University announced that Doyeon Kim, an undergraduate researcher in the Department of Electronic and Electrical Engineering (Advised by Professor Wansu Lim), has published a paper in a world-renowned academic journal. The research focuses on implementing drone Artificial Intelligence (AI) identification technology through low-power semiconductors. This study is drawing significant attention as it can dramatically improve the battery efficiency of systems that monitor unauthorized drones in real-time within military and industrial environments.

With the widespread use of drones globally in recent years, the technology to accurately detect and identify unauthorized drones has become crucial. Previously, AI technology was utilized to detect drones, but it suffered from a major drawback: it required an immense amount of computer calculations and consumed a vast amount of electricity. For this reason, it was highly challenging to directly embed AI into small, battery-powered surveillance devices or mobile equipment.

To solve this problem, Sungkyunkwan University student Doyeon Kim developed a new AI system named 'UAV-NAS.' Instead of the conventional method where humans manually design the internal structure of AI, this technology allows the computer to autonomously find the 'leanest and most efficient AI architecture' that can best analyze drone signals.

The research team designed this newly discovered AI to fit perfectly into a custom semiconductor chip (FPGA) that consumes very little power, and then conducted experiments. As a result, they succeeded in highly accurately identifying the types and flight status of drones in the sky while reducing power consumption by 88.7% compared to running the AI on a standard computer component (CPU). Because it operates intelligently while using very little electricity, it opens the door to monitoring drones 24/7 without worrying about battery life in outdoor or industrial fields.

This achievement was built upon the solid foundation laid by Doyeon Kim, who has consistently researched the semiconductor (FPGA) field in the university laboratory since his junior year. Notably, he was selected for an internship at 'Rebellions,' a leading AI semiconductor startup in South Korea, and never stopped his efforts to merge academic research at the university with practical experience in the corporate field. This study, which achieved the fruitful result of publishing in a world-renowned journal while excellently balancing undergraduate research and a corporate internship, is evaluated as a prime exemplary case demonstrating the global talent cultivation and 'Student Success' values pursued by Sungkyunkwan University.

Doyeon Kim, who led the research, will graduate with his bachelor's degree this August. Starting this coming September, he will pursue his Ph.D. program at the prestigious Purdue University in the United States on a full scholarship to continue his in-depth research on semiconductors and artificial intelligence.

Professor Wansu Lim of Sungkyunkwan University, the advising professor, said, "The achievement of student Doyeon Kim, who independently led world-class research and proved his practical competence externally since his undergraduate days, is a great pride of our university." He added, "We will continue to provide full support and educational environments so that future science talents, including middle and high school students, can freely pursue their dreams and leap onto the global stage at Sungkyunkwan University."
Angehängte Dokumente
  • ▲ Overall framework of the proposed technology. Radio (I/Q) signals from drones are fed into a neural architecture search (NAS) process that automatically designs a lightweight AI model, which is then deployed on a low-power FPGA chip for real-time drone identification.
  • ▲ Performance validation results (confusion matrices) showing accurate classification of drone types and flight modes across various conditions: (a) 13-class flight-mode identification, (b) 7-class drone-type classification, and (c) 23-class multi-dataset extension.
Regions: Asia, South Korea, North America, United States
Keywords: Science, Space Science

Disclaimer: AlphaGalileo is not responsible for the accuracy of content posted to AlphaGalileo by contributing institutions or for the use of any information through the AlphaGalileo system.

Referenzen

We have used AlphaGalileo since its foundation but frankly we need it more than ever now to ensure our research news is heard across Europe, Asia and North America. As one of the UK’s leading research universities we want to continue to work with other outstanding researchers in Europe. AlphaGalileo helps us to continue to bring our research story to them and the rest of the world.
Peter Dunn, Director of Press and Media Relations at the University of Warwick
AlphaGalileo has helped us more than double our reach at SciDev.Net. The service has enabled our journalists around the world to reach the mainstream media with articles about the impact of science on people in low- and middle-income countries, leading to big increases in the number of SciDev.Net articles that have been republished.
Ben Deighton, SciDevNet
AlphaGalileo is a great source of global research news. I use it regularly.
Robert Lee Hotz, LA Times

Wir arbeiten eng zusammen mit...


  • The Research Council of Norway
  • SciDevNet
  • Swiss National Science Foundation
  • iesResearch
Copyright 2026 by DNN Corp Terms Of Use Privacy Statement