Our aim is to provide a quantitative basis for improving and optimizing the non-visual, non-speech information displays used in blind-accessible instruments and devices. These multi-modal """"""""auditory/haptic"""""""" displays are used to provide feedback in a huge array of devices, instruments and travel aids via a combination of auditory cues and physical adjustment or manipulation (e.g., a knob, slider, or touch pad). Unlike synthetic speech, these displays tend to be used whenever a rapid adjustment, searching, scanning or aiming task has to be performed in real time. Examples include everything from accessible stud finders to the aiming cues for Talking Signs(R) receivers. Although auditory/haptic displays pervade the field of blind-accessible technology, no objective data exist on performance in terms of speed, accuracy or ease of use of the different display schemes or how to optimize performance within a given scheme. Designers usually adopt a display format by happenstance, based on past use or what is easiest to implement in their hardware. Our preliminary data demonstrates that speed and accuracy of such displays differ significantly, in a non-obvious way, demonstrating the need for objective information that could significantly ease the use and increase the efficiency of assistive technology and of instruments and appliances that utilize multi-modal outputs and/or universal design. To study the problem objectively, we categorize the most common audio-haptic displays as step, spike, trough and peak types. These classifications hold whether the auditory cue is pitch, volume, repetition rate or some other variable. For each of these display classes, we will investigate the dependence of performance on display characteristic (e.g., spike width, slope, etc.), then compare the relative effectiveness of the four display types at each display's point of maximum effectiveness. We will also extend these one- dimensional findings to explore two-dimensional haptic conditions, in two sets of experiments using the best-performing auditory display classes for both radial (with only one auditory feedback parameter) and x-y (with two auditory feedback parameters, one for each spatial component). The results will enable us to publish and disseminate concrete audio-haptic display design guidelines for researchers and industry, facilitating major improvements in speed and accuracy of future technology accessibility features for blind and visually impaired citizens.
The proposed research will compare several types of audio displays commonly used in guidance devices and electronic instruments for the blind. Although such audio feedback techniques are often used, little is known regarding which ones are most effective. The results of this research are likely to yield significantly-improved travel devices and other tools used by blind and visually-impaired people through an improved understanding of how to apply audio display techniques.
