Engineering Design and Psychophysical Evaluation of a Wearable Vibrotactile Display

Abstract

This thesis documents our research in high-fidelity, low-cost haptic interfaces and the objective means by which they can be evaluated. Our preliminary work evaluated the ability of a simple, shape-memory alloy, binary tactile feedback system to convey contact and grip force information to a user. We obtained counter-intuitive results: the binary touch feedback did not provide any benefit (according to objective performance measures) to the user while he/she performed a pick-and-place telemanipulation task with good visual feedback.

This finding led us to develop a wearable vibrotactile glove comprised of miniature voice coils that provides continuous, proportional force information to the user. The most effective method of conveying vibratory information to a user's fingertips was explored through several psychophysical studies. We discovered that by co-varying the frequency and voltage amplitude of the stimulators we are able to expand the user's perceptual response range compared to conventional methods of varying the frequency of a fixed-amplitude signal or varying the amplitude of a fixed-frequency signal.

After achieving successful subjective evaluations of our vibrotactile glove, we then demonstrated the objective utility of the glove by measuring the performance of many users (e.g., mean grip force, task completion time) while conducting several telemanipulation tasks. A force-limited pick-and-place manipulation task showed that the proportional tactile feedback allowed the users to control the grip forces more effectively than no feedback or binary feedback. This same task also allowed us to review and refine our hypotheses from the preliminary experiments and conclude that the very nature of the binary tactile feedback (it is binary) does not provide any perceptual advantage to the user. A sorting experiment established that proportional tactile feedback enhances the user's ability to discriminate the relative properties of objects, such as weight. A spring-loaded box insertion task illustrated that proportional tactile feedback allows the user to learn more difficult remote manipulation tasks more quickly. All three experiments provided statistically significant evidence of the usefulness of proportional vibrotactile feedback, compared to no tactile feedback.