Semi-Autonomous Car

The design project for this my Senior year was to construct a model car prototype which consists of two fourth generation "intelligent" microprocessors at Clarkson University. Initially placed off of a looped track, the model car will be capable of locating and traveling around the track. It is also capable of navigating autonomously and avoiding collisions with other objects on the track, and successfully navigating intersections around the loop. The original product concept is derived from the Freescale Cup. The Freescale rules used for the competition will serve as the guidelines for our product, but the Clarkson Intelligent Cars IV will have additional functional capabilities.

The intelligent car multiple vision sensors including a linescan camera and a CMUcam5 Pixy camera developed by Carnegie Mellon University. The linescan camera will be used primarily to follow the track. The Pixy has a resolution of 640 x 400 pixels, and it uses a hue-based colour filtering algorithm to detect coloured objects of several different signatures. The car will use the Pixy sensor to detect and navigate intersections using command signatures. These signatures provide instructions to the car to turn left, right, turn around or continue when it comes to an intersection on the track.

Below is the high level design of the system:

Google Docs link: https://drive.google.com/open?id=1i-51Vecy44v9vysUjhKCxUhc5ETamGdj