Developing autonomous vehicle research platform-a case study

Abstract

The article presents the case study of autonomous vehicle research platform development. All the important aspects of car conversion are presented. Firstly the actuation from manual to automatic acceleration, braking and steering is discussed. Secondly the Nvidia Drive PX 2 GPU based embedded platform chosen as the control unit as well as the ROS software are also described. Thirdly all the sensors that are needed to allow the car to sense the environment around (cameras, radar, LIDAR, precise GPS) and react to any dangerous situation are presented. Finally the vision algorithms that were developed based on the designed autonomous vehicle research platform are described.