



The Non-Line-of-Sight Leader Follower is a navigation system that allows follower vehicles to autonomously follow the path of a human driven leader vehicle. Robust position estimation for each vehicle is provided through a Kalman filter formulation that integrates data from inertial navigation sensors, odometry, GPS and ALVINN road following vision software developed at Carnegie Mellon University. The leader vehicle periodically transmits its position via radio link to the follower. The follower reconstructs these position points into the leader's path, thus allowing the follower to recreate that path while being out of sight of the leader.