Methodology for Translating Upper Extremity Motion to Haptic Interfaces

Abstract

The goal of our research is to design a haptic-based system to assist people with disabilities in learning functional motor tasks. Specifically, we are studying the upper extremity while performing daily activities. We have explored fine motor activities, such as writing, and gross motor activities, such as feeding and grooming. We use electromagnetic sensors (MotionStar Wireless 2) to capture unencumbered movements performed by a 'normal' individual. The captured movement is translated to the haptic coordinate system with the use of a body centered, intermediate frame for gross motor activities and with the use of a table-top centered frame for fine motor activities. We segment the movement into simple strokes and then 'smooth' the strokes using low pass filtering, cubic spline fitting and polynomial curve fitting. Our system varies the haptic force as the function of performance of the trainee. We have used our method to program the Phantom Omni for writing tasks and the Phantom Premium 3.0/6 DoF for gross motor tasks. We demonstrate our methodology with several examples of fine and gross functional activities.