ROBOTIC PLATFORM FOR AUTONOMOUS VEHICLE SAFETY ASSESSMENT

Abstract

Autonomous cars have risen in popularity and are beginning to appear more on our roadways. Currently there are no measures to monitor their behavior and their behavior has not been fully refined enough to allow them to operate without human intervention. Currently an artificial intelligence software has the ability to track vehicles visually through a traffic camera network, but there is no way to verify its accuracy. The proposed problem was put out by the Arizona Commerce Authority and the Institute of Automated Mobility and a solution was developed by Tucson Embedded Systems and a team of University of Arizona Students. The designed solution was a scaled RC car platform that featured a GPS module and inertial measuring unit that would collect positional and kinematic data while also having a visually recognizable checkerboard calibration pattern. This platform would allow for small scale lab environment testing, but would also allow for the platform to be placed in a road car with a scaled up calibration pattern for use with traffic camera systems on roadways. The student team was able to develop and deliver a working product that would allow for the integration of autonomous capabilities into the platform in the future.