Evaluation of terrain using LADAR data in urban environment for autonomous vehicles and its application in the DARPA urban challenge

This paper describes the autonomous ground vehicle developed by researchers at the University of Florida that participated in the 2007 DARPA Urban Challenge. Specifically, this paper introduces LADAR based terrain evaluation algorithms for an urban environments as well as off road. The terrain evaluation algorithm is very important for safe driving at high speed. On the real road, the driver is faced with numerous road conditions such as the smoothness of the road surface, curbs, or debris. For an unmanned vehicle to be successful, the algorithm has to decide whether the surface is traversable or non-traversable. For this reason, this paper focuses on the problem of extracting the ground terrain surface from 3-D point clouds obtained from LADAR sensors. The paper outlines the approach used by the University of Florida's Team Gator Nation to address the question of classifying traversable road conditions.