In this research, the way to support to design automated production line by making a correlation between human working and robot's based on analyzing of human motion was developed. It was seen assembling works as the combination of simple motions of work. The motion of work was became segmented by measuring acceleration and angular speed data and acknowledging motions of simple works and transient states them by using Support Vector Machine(SVM).The working within segmented motions was extracted by using Self-Organizing Maps (SOM) and made correlation between human workings and robot's functions. By the medium of them, the basis of the system to derive robot's features for automated production line was developed.