Ikhyeon Cho
Ph.D. Candidate
Expected Feb 2027
I am a Ph.D. candidate at Korea University, where I also received my B.S. in Mechanical Engineering. I am advised by Prof. Woojin Chung in the Intelligent Systems & Robotics Lab. During my Ph.D., I have led several industry-sponsored research projects, including collaborations with Hyundai Motor Group Robotics LAB and Hyundai Mobis.
I am interested in how robots turn perception and experience into intelligent action. I work on SLAM, spatial perception, and experience-driven robot learning, with self-supervised learning as a key approach. Building on my mobile robotics research, I aim to explore embodied intelligence—how robots develop and adapt task-driven behaviors through interaction with the physical world.
Research
My research aims to make robots more capable and reliable in the real world, with an emphasis on the following areas:
Learning from Robot Experience
Robotic Autonomy in the Real World
ODS-Bot, Robot-in-the-Loop Simulation,
Multi-Robot Navigation
Publications and preprints
Terrablox: Evidential Persistence Reasoning for Robot-Centric Elevation Mapping in Dynamic Urban Environments
IEEE Robotics and Automation Letters (RA-L)Under Review
MapGLoc: Robust LiDAR Global Localization on Dense 3D Maps for Mobile Robots
IEEE Robotics and Automation Letters (RA-L)Under Review
Learning Self-Supervised Traversability With Navigation Experiences of Mobile Robots: A Risk-Aware Self-Training Approach
IEEE Robotics and Automation Letters (RA-L), 2024
Open-Source Software
FastDEM Official Implementation
Real-time 2.5D elevation mapping engine for embedded mobile robots.
TraverseSet: A LiDAR-inertial dataset for traversability learning
LiDAR, IMU, and odometry data from urban and rural environments.
Selected Projects
- Project Lead
Learning Robot-Specific Traversability from Navigation Experience
Description
Predicts robot-specific traversability in real time by learning from short driving demonstrations with automatically generated labels.
- Project Lead
Fast Elevation Mapping for Terrestrial Robotic Exploration
Description
Provides real-time terrain maps for resource-constrained ground robots using LiDAR or RGB-D data.
- Project Lead
Multi-Robot Perception and Navigation in Dynamic Factory Environments
Description
Enables multi-robot navigation around unexpected obstacles through camera–LiDAR perception, shared obstacle tracking, and adaptive replanning.
- Project Co-Lead
Robot-in-the-Loop Navigation Testbed
Description
Connects physical robots with Isaac Sim for consistent evaluation of autonomous navigation across robot platforms and software stacks.