Human skill and brain activity in machine operation

This paper reports experimental analyses of human characteristic in machine operation by utilizing a stabilization task. For this aim, a test system composed of a force-feedback haptic device and a real-time computer graphics of a virtual pendulum was developed. Two kinds of analyses to the pendulum's movement and hand's motion revealed special characteristics of a skilled person: (1) a bode gain plot of the pendulum's movement becomes complex from one decline line to three decline lines, and (2) a spread at faster velocity range appears to a Gaussian distribution on velocity of human control. Additionally, a near-infrared spectroscopy (NIRS) was used to monitor cortical changes in haemoglobin at an area covering the primary motor/sensor cortex in brain. It could be confirmed that human who mastered the stabilization task sufficiently (a) utilizes sensory information of wide body part, (b) reinforces on active visual sensing, and (c) decreases a ratio of paying attention to motor control of one's hand.