I am currently a graduate student at the University of Toronto Institute for Aerospace Studies and Robotics Institute working in the Space and Terrestrial Autonomous Robotic Systems (STARS) Lab under supervision of Dr. Jonathan Kelly. My research work involves observability- and noise-aware state estimation and self-calibration for aerial and ground robots. The goal of this work is to increase performance, ease of use and robustness of robotic systems deployed in real-world environments. This website presents some of the projects that I have had the opportunity to work on. For any questions, feel free to reach out via email!
Below are some of the projects I have worked on…
Performed multi-UAV tracking based on lidar point cloud data through filter-based state estimation, data association and outlier rejection. lidar sensors provide the ability to effectively and spatially locate UAVs in areas such as airports where unaccounted for UAVs pose a significant risk to the general public. The above image shows collection of the mutli-UAV datset using an OPAL lidar.
Implemented a radar-camera EKF with data association and outlier rejection for the course AER 1513H State Estimation for Aerospace Vehicles. Project utilized a Continental ARS4-A 2D automotive radar and simulated camera data to estimate the position of a radar corner reflector. Ground truth data was obtained using a Vicon motion capture system. Additional results for this project can be seen via video or report.
Modeled, simulated and experimentally verified OPAL Lidar viability in various applications in addition to general performance capabilities. Publicly available information available in the paper Aerial and Surface Security Applications Using Lidar. The above image was taken during field testing.
Various perception work on the KITTI Dataset completed for the course AER 1515H Perception for Robotics. Work included stereo vision feature detection, matching and outlier rejection through epipolar geometry-based RANSAC (top image) as well as multi-vehicle tracking using lidar (bottom image).
Replaced the side mirror of a car with a low cost camera and Lidar sensor for an undergraduate project. A display mounted inside the car provides the driver with a mirror feed overlayed with detected vehicles and a colored warning system based on the speed of the passing vehicle.