Comprehending a hand motion language in shape conceptualization

This paper reports on a study concerning the comprehension of an experimental hand motion language in shape conceptualization. Hand motion is regarded as a prospective input mechanism for computer aided conceptual design systems for initial shape design of consumer durables. Our hand motion language has been developed based on the analysis of the (a) information necessary to describe shape concepts, (b) descriptive and indicative capabilities of human hands, and (c) cognitive and perceptive aspects of processing hand motions. This language was used in designed experiments to describe simple, compound and hybrid shapes for designers and tested with respect to comprehensibility. The subjects were asked to reconstruct the presented shapes by sketching on paper. Comprehension of the hand motion language has been evaluated in terms of fidelity and efficiency. Fidelity was expressed in terms of the number of deviations of the sketches of the test shapes from the existing CAD models. Efficiency was expressed as the proportion of the time needed to understand the test shapes communicated by hand motions and the time needed for the presentation. The results clearly indicate the potentials of a HML in shape conceptualization. In addition, the experiments revealed several new issues related to the application of the hand motion language in a multi-modal interface. The most important one is the need of chunking of the hand motion sequences in a way that enables the computer system to reliably reconstruct shapes and the designers to understand the formation of shapes on a higher semantic level. These issues will be addressed in our future research