Publication

Autonomy Artificial Intelligence Robotics (AAIR)
2009

DYNAMIC AUTONOMOUS GROUND VEHICLE RE-ROUTING IN AN URBAN ENVIRONMENT USING A PRIORI MAP INFORMATION AND LIDAR FUSION

by Christopher Mentzer; George McWilliam; Kristopher Kozak

Abstract

As part of an Internal Research and Design effort to take existing disparate technologies and integrate them into a single autonomous vehicle to advance the state-of-the-art in unmanned ground vehicle autonomy, SwRI has developed a data representation and routing algorithm to deal with the complexities of interconnecting urban roadways and the static and dynamic hazards in such an environment. The program was designed to utilize data from a Route Network Definition File (RNDF), which contains a priori roadway network data. Using its known location and a given destination, the vehicle determines the shortest route to completion. If, during traversal of that route, the vehicle detects an obstacle in its path using its on-board sensors, it will dynamically re-route its path whether that requires changing lanes on a multiple lane road or turning around completely and finding a different route if the path is completely blocked.