Skill analysis of human in machine operation

A purpose of this paper is to discover catholic characteristic of human skill in machine manipulation for realization of a Human Adaptive Mechatronics(HAM). HAM is a novel concept of an intelligent mechanical systems that adapt themselves to the user's skill and assist to improve the user's skill. This paper reports experimental analyses of human characteristic in machine operation by utilizing a stabilization task. By using a force-feedback haptic device and a real-time computer graphics of a virtual pendulum, processes of subjects to be skilled were observed, and three kinds of analyses were applied. The first is probability distribution analysis of motion, the second is monitoring of brain cortical by a near-infrared spectroscopy(NIRS), and the last is correlation analysis to estimate human control law. It could be confirmed that human who mastered the stabilization task sufficiently (i) utilizes sensory information of wide body part, (ii) reinforces on active visual sensing, and (iii)predicts velocity information by using dynamics model which is formed in brain.