Performance analysis of HMI system based on PCA

Abstract

This work proposes a performance analysis method based on Principal Component Analysis (PCA) for the skill quantification during several pursuit manual tracking tasks, which may occur in human-machine manual operations such as piloting, driving and navigation. A set of increased difficulty tracking experiments was made and analyzed using an 8-bit analog joystick human-machine interface (HMI). This PCA methodology based on the projections of the observation data which optimally captures the data variations, has confirmed to differentiate better the computed human-machine performance of the common operator/task, and can be used as a precision skill observer for a human machine assist-controller electronic device.