Trajectory Planning for Functional Wrist Movements in an ADL-Oriented, Robot-Assisted Therapy Environment

Abstract

A task-oriented robot-assisted therapy environment to support activities of daily living tasks was developed with the goals of addressing some of the shortcomings of current robotic therapy systems. An important aspect of making these environments work is the implementation of trajectory planning algorithms that support the naturally curving wrist movement seen in real life functional tasks. We explore the challenge of naturally supporting the positioning of the wrist for activities of daily living tasks such as drinking and feeding. In this paper, we examine the minimum jerk model often used to define trajectory planning routines to automatically position the wrist and present the results of fitting these two models to natural wrist movements for a drinking task. Also, we present a case study to examine how an able-bodied experienced the two models as implemented on the ADLER, task-oriented robot-assisted therapy environment.