Development of Navigation System for Autonomous Vehicle to Meet the DARPA Urban Grand Challenge

Spring Light team is the first, and the only, Asian team registered to the 2007 DARPA urban grand challenge. In this paper, we present the navigation system of the daughter dragon vehicle of the spring light team. The purpose of the navigation is two fold. First of all, it takes the road map, the mission description file and in-vehicle GPS data as input, and plans in real-time an optimized route connecting the vehicle´s actual position to the destination point as output. The optimization criteria considered in this work include the shortest traveling time, the shortest traveling distance, and the minimum number of traversed junctions. Second, the navigation system receives the data from visual perception such as laser-range finder, and generates a trajectory in terms of the vehicle´s steering angle profile and velocity profile as output. In this paper, we detail the algorithms implemented inside the proposed navigation system. And, the proposed navigation system has been fully tested with simulations. Real experiments are actually undertaking, and will be presented in the conference.