There is a graphic tablet even available by the German companies
Metek and Papenmeier, and they sold three specimen as far as I know.
It employs a matrix of 119 by 59 tactile pins driven
electromagnetically. Unfortunately, it has two disadvantages:
The resolution is 10 dots per inch. So, Braille is displayed well,
but lines are felt as a string of dots and the quality is far from
that of tactile paper graphics.
The price is in the range of $50.000.
These points address the main problems faced by each approach to
construct a tactile graphic output device:
With technologies used in state of the art tactile devices we are far
from a sufficient resolution. Indeed the absolute accuracy of the
tactual sensors in determining the position of a single sensation is
in the range of 2 mm, preventing the use of Braille characters being
smaller than those commonly used. But nevertheless one can distinguish
well between tactile lines with differences in height or in width of
less than 0.5 mm. While tactile graphics on paper or plastic take
advantage of different kinds of lines and surfaces we will be very
happy when displaying just one kind of a nearly continuous line on a
tactile display for the first time.
A convenient tactile graphic display should be read using both hands
and should not be smaller than the area covered by two spreaded
hands. A resolution of 20 dots per inch and an area of 16 by 12 inch
yield 320 by 240 dots, i.e. nearly 80.000 tactile display elements.
Given a price of $20.000 as a goal, a single display element is
limited to a price of 25 cents neglecting the price of the display
controller, the housing etc. But the price of one pin of a high
quality piezoelectric Braille display is about $10 today.
These problems can be overcome only be employing new technologies. I
know of two current approaches in this direction:
Prof. Steven Leeb at the Massachusetts Institute of Technology is
using polymer gels in order to construct a tactile tablet. I am
looking forward to meet him at the CSUN conference in Los Angeles on
March 94 for the first time, so unfortunately I am not able to
discuss his work today.
The second project is conducted by us at the Fernuniversitaet Hagen,
Germany. We are using electrorheological fluids in order to lift and
to settle little "pimples" in a membran covering the surface of the
tablet, and we try to overcome the price problem by integrating the
display elements in the board during the fabrication processes. For
example, due to integration, a 386 processor is far less expensive
than the equivalent number of discrete transistors. The display will
be an input tablet as well, so the displayed objects can be
manipulated in a very direct way. We expect to have some little
specimen in 1994 and a large prototype in 1995.
I am not sure if I should burden the capacity of the net by
distributing an article concerning this work to all subscribers of
this group, but if someone out there is interested in a more detailed
description I could send it to his or her email address directly.
Following the discussions in this list on GUIs and on the usefulness
of graphics for blind people is very interesting for me, because I
myself am a sighted computer scientist (with no preference for
GUIs), and I am aware of the engineering problem of doing what is
possible instead of what is needed. Of course, I asked many blind
people for their opinion before starting our project but even they
had no experiences with graphic displays, obviously. I can imagine
how to use a graphic display in some domains of education and
research, when reading texts illustrated with drawings, and while
reading mobility maps, for example. I am not sure if GUIs should be
presented in a moderate graphic form showing text items distributed
on the screen or if they should be transformed to a pure text line
based interface. Probably it depends on the user. Surely, it is not
useful to display things like icons because they are much harder to
recognize than text and the graphic presentation delivers no
Have a merry christmas,
This archive was generated by hypermail 2b29 : Sun Dec 02 2012 - 01:30:03 PST