Automated Route Finding on Digital Terrains

Path planning is an important problem in Artificial Intelligence and Robotics. In this paper we examine the problem of finding an optimal path for a road connecting two points in a digital approximation of a real terrain. Additional constraints, such as terrain gradients, will be introduced in order to achieve more realistic roads. A performance comparison will be made between the standard A* algorithm and variations of it. Different heuristics, which are used by the algorithms to guide them to the goal node, are presented and compared. To overcome some of the computational constraints associated with route finding on large digital terrains we introduce multi-resolution searching. We will show that multi-resolution searching greatly reduces the time associated with the route finding process while still resulting in near optimal solutions.