This paper describes a path planning and a consecutive operation based on discrete acquired motion data for an auto parking system. In order to park a car in various situations, it is necessary to plan the path that contains the turn back motion according to the situation. However, it is difficult to search the appropriate path because there are many paths to reach the target position. Therefore, the method by which combination of motions is decreased by limiting to discrete motions was proposed. Moreover, a method that generates a consecutive operation from discrete motion data was proposed. The proposed methods were applied to a model car of 1/6 scales. The results of simulations and experiments in a real environment show the effectiveness of the proposed methods.