Display of Virtual Braille Dots by Lateral Skin Deformation: Feasibility Study

6. Conclusion and Future Work

This study showed that experienced Braille readers could read sequences of first-row Braille characters using the VBD with a legibility ranging from passable to excellent. This is encouraging considering that most subjects had little prior training with the device and that the character strings were meaningless.

Reading with the VBD was nevertheless difficult. The control experiment showed that subjects could read faster and with perfect accuracy when the strings were presented on embossed vinyl tape. This observation is confirmed by the subjects' verbal comments during both experiments.

Adding texture to the dots was found to increase performance for some subjects but was rejected by others. Prolonged use also seemed to cause tactile fatigue (numbness in the reading finger) as reported by subjects for both nominal and textured stimulus. This phenomenon was sometimes confirmed by an increase in the number of reading errors over time in the case of textured stimulus.

While this study was conducted with too few subjects to make it possible to draw final conclusions, it suggests nevertheless that reading Braille characters with devices based on the principle of the VBD could be possible. The study also helped identify the strengths and weaknesses of the current prototype and, more importantly, provides indications as to how it could be improved.

The strength of the dot sensation must be increased to realistically convey the illusion of a Braille dot. This issue involves the improvement of the actuators used and of their configuration. On-going work concerned with the micro-mechanical properties of the skin and the means to deform it at a very small scale is expected to yield an improvement in the performance of piezoelectric benders for this application [Wang et al. 2004]. Further experimentation with deflection functions could also lead to better approximations of the sensation of scanning over Braille dots. Designing deflection functions for maximum strain variations at skin mechanoreceptors may, for example, increase the strength of the dot sensation. Indeed static mechanical models of the skin designed by Phillips and Johnson [1981] and Van Doren [1989] suggest that deformation of receptors, as opposed to their displacement, is likely to determine sensation. The strain experienced by slowly adapting (SA) receptors may be particularly important to resolve the spatial details of scanned Braille [Phillips and Johnson 1985].

The cause of tactile fatigue must also be addressed. It is not clear what causes it and how it can be avoided. It is likely however that replacing the contact line by a more uniform contact surface could significantly delay the onset of tactile numbness.

Reading with the VBD requires a scanning movement. While this allowed the VBD to render adequately the dynamic sensation of sliding over dots, it prevented the user from stopping over a region as is sometimes done when reading physical Braille. Display without net movement could be possible if the magnitude of the strain produced by the device was made to depend on the force applied by the subject. This not only could make the sensation of sliding over a dot more realistic, but also allow the subject to stop and press against virtual dots.

The results of the tuning experiment and of the legibility experiment conspire to indicate that perhaps the greatest problem with the current display design is the difficulties experienced by subjects in evaluating the distance between dots. This is suggested by the distortions in perceived Braille dimensions introduced by the VBD when compared to standard English Braille. This is also consistent with observed reading errors and verbal reports of the legibility experiment.

Finally, the VBD should be extended to display complete Braille characters. Packing 4 rows of actuators capable of displaying forces and displacements similar to those of the VBD within the height of a Braille cell will be a significant technical challenge.