EVALUATION OF MOTION MAPPINGS FROM A HAPTIC DEVICE TO AN INDUSTRIAL ROBOT FOR EFFECTIVE MASTER-SLAVE MANIPULATION

摘要

Master-slave systems with identical master and slave arms have been in vogue for many years now. With the advent of computer-assisted master-slave manipulation technology, it is now convenient to use a small haptic device as a master, while a standard industrial robot serves as a slave. However, the challenge in this case is to select a suitable motion mapping from the haptic master to the slave robot. The problem is not trivial, as their degrees of freedom, workspace and inertia are all widely different. Various kinds of motion mapping have been suggested in the literature for a haptic master. We have devised a new mapping called boundary drift control. We present here the result of an experimental evaluation of the effectiveness of some of these mappings, including the one devised by us. A series of tests were conducted with a select group of operators to evaluate and compare the mappings in terms of efficiency and accuracy. Statistical significance of the observed data is established through ANOVA analysis. The role of skill, if any, in this evaluation is explored through another set of experiments. Qualitative feedbacks from the operators about ease of use of these mappings are also recorded. This study provides an insight into how to select a suitable motion mapping for a haptic master for a given job. We found that boundary drift control is well suited when both speed and accuracy are on demand.