Autonomous Vehicle Navigation Testbed

The team was asked to design an omnidirectional vehicle test bed and laboratory to enable faster and more cost-effective testing of autonomous vehicle navigation algorithms. Lab users can load algorithms into a microcomputer mounted on the vehicle, which is fitted with reflectors that allow infrared cameras to track its location. Algorithmic efficiency can be determined by comparing data from cameras with data from algorithms. An Xbox Kinect, inertial measurement unit, and six ultrasonic sensors provide information for the algorithm. Navigation calculations are performed on a Raspberry Pi 2, and an Arduino Zero is used to interface with sensors. Movable obstacles are placed in the testing area to allow users to custom build test scenarios. The test bed and laboratory have been designed with future expansion in mind. Subsequent projects could involve simultaneous use of multiple vehicles. The omnidirectional capability of the vehicle allows land- and space-based vehicle navigation algorithms to be tested.