A Motion Verb Interface to a Task Animation System

TEMPUS, an interactive software system designed at the University of Pennsylvania, provides a graphical simulation of human movements within a workstation environment [4, 5, 17). The existing user interfaces to TEMPUS are based on menu selection, alphanumeric commands or data input and graphical interaction. The latter are especially suited to the types of interaction necessary for creating workstation objects and performing body positioning in TEMPUS. Looking toward the future application of TEMPUS, however, we see that the long-term goals of human motion modeling will include the analysis and visualization of extensive tasks in space involving one or more individuals working in concert over a period of time. In this context, the TEMPUS body positioning capability, though extremely useful in creating and validating a small number of particular body positions, will become somewhat tedious to use. The macro facility helps somewhat, since frequently used positions may be easily applied by executing a stored macro. The difference between body positioning and task execution, though subtle, is important. In the case of task execution, the important information at the user's level is what actions are to be performed rather than how the actions are performed. Viewed slightly differently, the what is constant over a set of individuals though the how may vary.