6. The robot controller as set forth in claim 4, wherein the learning control part searches a learned motion from past learning data, before executing a regenerated motion
based on the learned motion speed-up ratio, and then, when the learned motion similar to the regenerated motion exists in
the past learning data, the learning control part corrects the motion speed-up ratio of the regenerated motion based on and
the motion speed-up ratio associated with the learned motion and a difference between motion information of the regenerated
motion and the learned motion.