Visual-Lidar Odometry and Mapping (VLOAM) payload
On Work Term 2, developed a visual-lidar odometry and mapping (VLOAM) payload for a Bell 412 helicopter to enhance its autonomous capabilities for the Memorial University Intelligent Systems Laboratory and The National Research Council.
- Implemented and debugged an IMU, LiDAR, Thermal Camera, RGB Camera, and Radar on a Nvidia Jetson with Linux OS, ROS and GitHub for data visualization and management.
- Synchronized data through trigger-signals and timestamps using communication protocols, Arduino, circuit analysis and ROS to improve payload accuracy in GPS-Denied operation.
- Tested sensor fusion using state-of-the-art localization and mapping packages in a variety of conditions for research data collection.
- Awarded the NSERC Undergraduate Student Research Award
Final Design of VLOAM Payload
Synchronization Flow Diagram
LOAM mapping of sea ice in St. Johns, Newfoundland harbour. Data used to study autonomous flight over sea ice.
Flat Design of Payload
LiDAR drift in the lab, needed to fix intrinsic calibration of IMU
Bell 412 Helicopter
Basement of the Engineering building map before LiDAR fusion with IMU
