Autonomous path planning in a variety of environments

This paper presents the path planning component of a control architecture for autonomous vehicles. Path planning is accomplished by using geometric formulas to calculate straight line segments that represent each stage of the overall path. The line segments are compared to a list of known obstacles in the environment. If collisions are detected, new path segments are generated to re-route the path around the obstacle. Unsafe areas are modeled as obstacles in this algorithm. This paper presents the STESCA control architecture for autonomous vehicles and then describes the path planning algorithm and data structures, demonstrating the algorithm through a few examples.