The 2011 term project for MIE444 - Mechatronics Principles at the University of Toronto required the conception, design and prototyping of a line following vehicle that would traverse a maze of black lines set on a white background. Time trials were held between approximately 10 teams of 3 members each over a series of 3 runs of the maze to determine the fastest vehicle.
Each run involved a different starting point as well as different placement of the two wooden-arch "checkpoints" around the maze. Teams were required to pass both checkpoints, signal once they had passed a checkpoint and only then return to the starting point, at which point the time would be stopped.
This video shows the second run of the winning team, Team #2 (Hasanka Attanapola, Andrew Oldham and Sagar Saxena). This line-follower was programmed to turn left (when possible) at each junction. A red LED signals traversal of a checkpoint. A smart-algorithm was added to the code (not demonstrated in this video), to have the vehicle turn around and switch to a right-handed turning rule if both checkpoints were encountered prior to the vehicle traversing half of the maze. This provided significant advantage if checkpoints were both close together and near the starting point as the vehicle would not have to traverse the entire maze, rather, it would pass both checkpoints and then turn around and retrace its steps back to the starting point.
The Quanser QIC board was used to to control the system, taking analogue readings from five infrared sensors placed on the bottom of the vehicle in a forward V-shaped pattern. This arrangement and spacing of the sensors provided adequate resolution when detecting the black line and resulted in very precise movement when traversing the maze, as is evident in the video. This allowed for the vehicle to travel at a high speed without worry of overshoot or loosing the line. In fact the two 6V drive motors were run at 7.2V to obtain the edge in speed.