During round 2 of the SAE AutoDrive Challenge, I led the aUToronto Perception Object Tracking team, contributing to aUToronto's first-place win in the competition. My team developed the Object Tracking and Sensor Fusion software pipeline, which was used to secure first place in the Dynamic Obstacle Challenge.

I served as the head developer, leading a team of 5 students in writing a C++ ROS2 package used to fuse incoming detections and track objects over time. I assigned Jira tickets, conducted code reviews, and mentored the team while also actively developing code. The software architecture was designed by myself and the team captain, handling real-time tasks processing objects at 10 Hz. I developed the multithreaded sensor fusion message handling system, which queued detection messages from 11 sensors in different modalities (solid-state LiDARs, cameras, and radars) and grouped associated detections together. I also managed the data flow throughout the pipeline and performed integration of components.

Finally, at the competition, I presented aUToronto's entire software architecture in the Concept Design Event, where we achieved first place.