Autonomous Maze-Navigating Robot

Details

• Co-developed an autonomous maze robot in a 4-person team, combining onboard vision, embedded motor control, and inter-device communication, delivering the second-fastest final run in the class
• Built the Python/OpenCV perception stack using ArUco marker tracking to estimate heading and distance in real time, giving the robot reliable position cues for turn timing and corridor alignment
• Designed a multithreaded 640x480 vision pipeline with HSV-based color segmentation so image capture, marker detection, and navigation decisions could run concurrently without blocking control updates
• Streamed distance and angle telemetry from the vision subsystem to the team's Arduino motor controller over I2C
• Owned the team's Git repository, managing branches and integration across all four members' subsystems