Copyright (c) 1991 by David Lunney. This document may be copied and distributed freely so long as it is reproduced in whole and contains this notice.
By David Lunney
But if in addition to adapted human interfaces the computer is fitted with suitable interconnections to the physical world, a new universe of possibilities opens up: an adapted computer with data acquisition capabilities can make the laboratory accessible as well as the office. Most laboratory measurements are now done with instruments, and access to instrumental measurements can help to make scientific and technical careers more accessible to people with disabilities. The computer can acquire data from instruments and sensors, control experiments, and assist in the analysis of data. An adapted computer with a data acquisition system does not solve all the problems for every disability, of course. A person who cannot perform manual operations may still need some help from an assistant. And, as we shall see later, a visually impaired person may need data analysis tools that can substitute for visual graphs. In short, the adapted data ac- quisition computer cannot remove all the barriers encountered by disabled people in a laboratory, but it can lower a lot of them.
The ULTRA can also be adapted for other disabilities: for example, we added a voice recognition terminal to an early version of the system so that it could be used by students with upper limb disabilities, and we have written software for some voice-controlled laboratory experiments and a voice-controlled calculator (5). The current version of the system could also be adapted for breath control by adding an STD pressure sensor board and writing appropriate software.
After we had worked with the ULTRA for a while we found that although it could give visually impaired students better access than ever before to instrumental measurements, there were sit- uations in which the experimental data were too complex to be presented conveniently as spoken numbers or as rising and falling pitches. We needed a means of presentation which would enable visually impaired students to perceive patterns in data, and to get the kind of quick overview that a sighted person gets from looking at a graph. Taking our inspiration from a landmark art- icle by E. S. Yeung on auditory presentation of multivariate data (6), we set about developing schemes for representing complex data patterns as complex sound patterns. We have now developed a few effective methods for presenting multivariate data -- especially data from infrared spectra -- as recognizable and fairly memorable musical patterns, using a computer-controlled music synthesizer.
We are continuing and extending our work on auditory pre- sentation of data under a grant from the National Science Foundation. We are again using infrared spectra simply because they are a convenient source of information-rich data. In our present work we are using artificial neural networks to extract information on molecular structure from the spectra; this ap- proach gives results that are roughly comparable to those ob- tained from an expert system. Our next step will be development of schemes to map the network's output vectors into auditory parameters. The use of neural networks gives us a great deal of flexibility because a suitable network can be "trained" to map any input vector into any output vector, and the output vector can then be used to control a variety of auditory parameters. Our final scheme therefore will not be limited to infrared spec- tra, but will be capable of mapping any sort of multivariate data into sound patterns. (Readers who are interested in auditory presentation of data should consult Ref. 7 for an excellent review of the subject. Ref. 8 gives an overview of our approach to auditory presentation of data, including some things that we tried that didn't work.)
We have not been able to find a manufacturer who would be willing to build the STD version of the ULTRA system for a number of reasons. We intend to solve part of that problem by porting some of the ULTRA's software to the ubiquitous IBM PC and its clones. At the time we designed the STD version of the ULTRA very few serious data acquisition products were available for the PC, while about 2000 board level products were available for the STD bus. Now, the situation is reversed. The STD bus is in decline, and the IBM AT bus has become the most widely used bus for industrial data acquisition and control systems. It now would be possible to replicate all the ULTRA's important func- tions on an AT platform.
Also, IBM has introduced the Personal Science Laboratory (PSL), a flexible, expansible, modular data acquisition system intended for school laboratories. The PSL can accommodate several kinds of sensors, and is simple to program because it communicates with its host computer through a serial port. In educational laboratories many of the functions of the ULTRA could be performed by a PSL and PC system equipped with suitable speech output.
2. D. Lunney, R. C. Morrison, M. M. Cetera, R. V. Hartness, R. T. Mills, A. D. Salt, and D. C. Sowell, IEEE Micro, 3 (4), 19 (1983).
3. R. C. Morrison and D. Lunney, Journal of Visual Impairment and Blindness, 78, 418 (1984).
4. R. C. Morrison, et. al., Personal Computers in Chemistry, P. Lykos, ed., 164-176, Wiley-Interscience, New York, 1981.
5. R. C. Morrison, et. al., Journal of Chemical Information and Computer Science, 24, 271 (1984).
6. E. S. Yeung, Analytical Chemistry, 52, 1120 (1980).
7. S. P. Frysinger, "Applied Research in Auditory Data
Representation," Extracting Meaning from Complex Data: Processing, Display, Interaction, Edward J. Farrell, Editor, Proc. SPIE 1259, 130 (1990).
8. D. Lunney and R. C. Morrison, "Auditory Presentation of Experimental Data ," Extracting Meaning from Complex Data: Processing, Display, Interaction, Edward J. Farrell, Editor, Proc. SPIE 1259, 140 (1990).
Department of Chemistry and
Science Institute for the Disabled
East Carolina University
Greenville, NC 27858 USA