AV

A Study on Trustworthy Performance Index in Reinforcement Learning for Controllers using Customer Feedback

A methodology study that corrects the errors inherent in customer feedback and removes unnecessary signals, thereby improving the reliability of the Performance Index (PI) model and integrating it into the controller.

Hyundai · Control · Estimation

Robotics

Data-efficient Learning Framework for Vision-based Robot Manipulation Control

Building learning frameworks that achieve robust vision-based manipulation control from limited real-world data.

NRF · Vision-based manipulation

Robotics

Spatial Intelligence for Robotic Autonomy: From World Reasoning to Precision Control

We develop spatial intelligence for robotic autonomy by combining symbolic world models for spatial reasoning and task planning with geometry-aware, data-efficient robot policy learning.

NRF · Spatial Intelligence · VLA · Agentic AI

Robotics

Development of XR Copilot Technology Based on Non-verbal Behavior for Quality and Safety Assurance in Construction Component Fabrication and Installation

Developing XR copilot technology that uses non-verbal behavior to ensure quality and safety when fabricating and installing construction components.

IITP · XR · LLM · Non-verbal Behavior · Multimodal Learning

View completed projects on the faculty profile →