Motion planning
Computing a safe, feasible path or trajectory for a robot or vehicle through obstacles and traffic rules, often combining maps, prediction, and optimization.
Motion planning chooses how a robot or autonomous vehicle should move from its current state to a goal while respecting obstacles, dynamics, and rules. Planners often search a graph of feasible poses, smooth a path, or optimize a cost function that penalizes collisions, jerky steering, and rule violations.
In the 2007 DARPA Urban Challenge, winning teams paired perception (LIDAR, cameras, sensor fusion) with planners that handled merges, parking, and intersections. Classical planning remains central in robotics even when machine learning models predict how other agents behave.