RESEARCH 01
LiDAR–Camera Fusion for Road Segmentation
Semantic road segmentation using RGB camera data, LiDAR-derived depth information, modified SegNet, and early/late sensor-fusion architectures.
Graduate research, robotics development, experimental systems, and instructional work at the University of Waterloo.
Research spanning perception, sensor fusion, navigation, planning, and mobile-robot control.
Semantic road segmentation using RGB camera data, LiDAR-derived depth information, modified SegNet, and early/late sensor-fusion architectures.
Experimental path-following implementation using a holonomic Summit mobile robot, model predictive control, and Vicon-based pose feedback.
System-level robotics work combining perception, localization, navigation, control, and physical robot operation.
System-level integration combining Summit XLS navigation, multi-floor mapping, vision-based pose estimation, WAM manipulation, and elevator interaction.
TurtleBot navigation demonstrations created for robotics instruction, covering narrow passages, obstacle avoidance, global replanning, and dynamic-obstacle interaction.
Experimental support for path-planning research led by Dr. Jeong-woo Han, evaluating path choice under robot motion and achievable-speed constraints using a Summit mobile robot.
Virtual robotics laboratory development, technical documentation, verification, grading, and student troubleshooting for senior undergraduate and graduate robotics instruction.
Development and validation of a VMware-based robotics environment integrating Ubuntu, ROS, Gazebo, MATLAB, TurtleBot simulation, and ROS networking.
ROS fundamentals, TurtleBot simulation, RViz, sensor visualization, mapping, SLAM, localization, and autonomous navigation.
Particle-filter localization, motion and sensor models, parameter sensitivity, pose estimation, covariance, and computation-time analysis.
Sampling-based path planning, obstacle inflation, robot motion constraints, path optimization, and Pure Pursuit trajectory tracking.