DARPA Grand Challenge robots are a good example of robots that need to plan a navigational path over uneven outdoor terrain. Most mobile robots face planning challenges when operating outdoors. A new research paper from the USC Center for Robotics and Embedded Systems lab titled, Time Optimal Path Planning on Outdoor Terrain for Mobile Robots under Dynamic Constraints (PDF format), present one solution to this problem. The authors, Marin Kobilarov and Gaurav S. Sukhatme, used the Segway Robotic Mobile Platform to develop and test a planner that uses "bidirectional rapidly exploring random trees (RRT) on the terrain surface with nodes connected by continous curvature trajectories". Basically, the algorithm allows the robot to plot the shortest path between two points over uneven terrain. More info and some nice graphics can be found on Marin's website.