Task and motion planning describes a class of AI problems where a robot has to reason about solving long-horizon interactions with the world through actions and motions. A solution to this problem discovers a sequence of feasible actions, each of which is enacted by the robot using a feasible continuous motion.