Autonomous driving research involves cooperation between engineers across fields, from control systems to machine learning. However, currently no commercial tool provides a full-stack solution for a self-driving vehicle development. OSCAR (Open-Source Self-Driving Car) aims to provide a fully functional autonomous driving platform and easy-to-use tools to speed up the autonomous driving development process
Conceptual overview of OSCAR
A common autonomous driving platform consists of a target vehicle, a sensor suit, and a processing unit, such as NVIDIA embedded board, running the Linux operating system with the Robot Operating System (ROS) middleware. For a developer with insufficient experience in programming, working with such platforms is a big challenge. Through OSCAR, developers can easily implement algorithms for end-to-end deep learning systems, computer vision-based and LIDAR-based systems using an end-to-end design tool, OSCAR IDE. OSCAR also includes a scaled autonomous platform and a remote laboratory that allows developers to obtain reliable results at low cost via field testing.