A machine learning device of a controller observes data on a movement speed of each motor of a robot and an adjustment amount of the movement speed, a target speed of a tip end of the robot, and a movement path proximate to the tip end of the robot, as state variables expressing a current state of an environment, and acquires determination data indicating an appropriateness determination result of the movement speed of the tip end of the robot. Then, the machine learning device learns the target speed data, the movement speed data, and the movement path data in association with the adjustment amount of the movement speed of each of the motors of the robot by using the observed state variables and the acquired determination data.