Research

Research Highlights

Prof. Myeong-hyun’s research team achieves double championship in the International Conference on Robotics and Automation SLAM Challenge and receives the Best Paper Award.

영문 64869fbcd7f18

 

– From May 29th to June 2nd, at the 2023 International Conference on Robotics and Automation held in London, the team competed in the competition that involved the localization and mapping (SLAM) technology. They achieved first place overall as well as in the vision field.
– They proposed a solution that enables real-time autonomous robot navigation through the estimation of the surrounding environment’s navigable area and object recognition, and received the Best Paper Award.
 
The “Urban Robotics Lab @ KAIST” or “URL @ KAIST,” research teams led by Professor Myung hyun, excelled at the HILTI SLAM Challenge held during the 2023 International Conference on Robotics and Automation (IEEE International Conference on Robotics and Automation, ICRA) in London, United Kingdom from May 29th to June 2nd.
This conference, with over 7,000 participants, is the largest academic event in the field of robotics.

The HILTI SLAM Challenge 2023 was part of the Future of Construction Workshop program, organized by HILTI from the Principality of Liechtenstein, the Oxford Robotics Institute at the University of Oxford, and the Robotics and Perception Group at ETH Zurich.

The competition aimed to develop robust SLAM algorithms that can accurately map challenging environments where conventional SLAM algorithms struggle due to limited construction features, narrow indoor spaces, and low-light conditions.
A total of 80 teams from prestigious international companies and research teams participated in this competition, which has become one of the renowned competitions within ICRA since its establishment in 2021.

The research team achieved first place in the LiDAR session among a total of 63 teams by utilizing their self-developed Adaptive LiDAR-Inertial Odometry (AdaLIO) algorithm and various optimization frameworks (Pose Graph Optimization). In the vision field, they won first place in the academic category (second place overall) by employing their robust vision-inertial odometry algorithm (UV-SLAM) based on line features they developed.

They are also expected to receive a prize of 3,000 CHF (Swiss Francs) and 1,000 CHF, respectively.
In the previous year, the research team participated for the first time and received the second place in the academic category (fourth place overall).
Professor Myung hyun expressed his thoughts on the achievement, stating, “This was a globally recognized opportunity for our self-developed SLAM technology, and I believe it will contribute to the advancement of the robotics industry by utilizing it in various autonomous driving, walking, and flying applications.”
 
 
 
2 1 2