The Matrix of Confusion—a Classification of Robot Movement

Abstract

The human-machine interface for robots has received only limited attention as robots have developed. Most research has focused on the design of teach pendants, hand-held devices for programming and manual motion control. Results from these studies have been generally inconclusive as to the best control design for teach pendants. In a fresh approach to the area, the human-robot interface has been analysed. This has resulted in the development of a method of classifying robot movements called the ‘matrix of confusion’. The classification shows the robot motions an operator would see when using a given control for a unique combination of operator position, robot position, programming mode and robot configuration. The use of the matrix has helped to highlight the most important factors in the task of manual motion control of the robot. This has helped in the development of a new motion system for a PUMA robot which is currently being tested in comparative trials.