Development and pilot testing for virtual manufacturing tools with intelligent attributes

Abstract

The use of synthetic and surrogate tools in training robots and planning events within virtual environments has tremendous potential in making the programming of complex machines simple and easy. For example, by designating tasks off-line with the virtual tools, attributes such as process planning and collision avoidance can be automatically and efficiently incorporated. In our research we have experimented with the use of attribute laden virtual tools in various tasks such as robotic-based grinding and welding processes. Such attribute laden virtual tools aid human operators in path planning as well as in making decisions about the process itself. In this paper we have tested our concepts of virtual tools and the use of attributes such as physical, reflex and command actions. Four sets of experiments were conducted with human subjects. Two kinds of virtual tools were used in these experiments, one with guide plane attributes and the other without them. One set of experiments using the head mounted display interface, tested human performance and evaluated the effect of the learning process in using the virtual tools. Results showed that there was marked improvement in task execution time using the tools laden with guide plane attributes over the unencumbered virtual tools. This paper discusses various future applications of these virtual tools in manufacturing. It is observed that unlike non-haptic visual interfaces, where no physical feed back is available to the user, attribute laden tools could provide a much easier interface to robots.