Discontinuity Detection Algorithm for Three-Dimensional Trajectory Data Analysis in Telerobotics

Abstract

Discontinuities in human movement can provide information about the control law that an operator is using. Such discontinuities are revealed by change points in one or more of the trajectory’s spatial derivatives. They are indicative of abrupt variations in the movement and provide information about spatial sampling and error detection. These discontinuities can arise in telerobotic applications, for example, when human operators pause while waiting for time-delayed teleoperation data to arrive or while adapting to account for misalignments between a remote camera and the operator’s local manual input reference frame. This paper presents a method for detecting discontinuities from trajectory records during three-dimensional target acquisition movements made by a simulated telerobot operator when display and control coordinates are misaligned. We develop the geometric detection algorithm and illustrate its use on actual data.