I come from a Work Group of the Blind and Visually Disabled Union of Saxony. We are dealing with tactile graphics, with design and production of tactile pictures and with access to them.
First of all, I would like to show you that we really are able to produce such tactile pictures and I will give you some examples. The first one is the picture of a special street crossing in Los Angeles, the stack crossing. We found it in a tourist guide, designed a graphic for blind people from it and made it tactile. Below this stack highway crossing of Los Angeles you will look at the picture of the Golden Gate Bridge also in tactile manner. Of course, these are examples where the blind need help from sighted people for explanations. Another examples are the skyline of Manhattan and some pictures from my country, bridges, monuments, and so on. You can look at it during the break. These are examples of our tactile picture postcards.
In addition to this, we made, also, maps. Let me give you one from South America, also made in swell paper. The next one is a railway station and its environment. The aim of that one is to make clear how you have to go from the railway station to the street car stop, which is important for blind people. The last picture which I want to show you now is the outline of the body of a human being with some organs. It is made from swell paper too without any braille texts, which I would like to emphasize. Nevertheless it is used in a course of anatomy. Also, the environment of the railway station is without any text. Please let me explain later why we did so.
Usually we make tactile pictures using the swell paper. The reason is that there is a very short way from the computer to the tactile graphic. Another possibility is to use a plotter equipped with an embossing device. Then you can use very cheap foils. An example is this map of Germany here. At last I will show some examples drawn by means of a plotter using special swelling inks. This one is a dog, a very simple shape. I've also taken with me a foil from this course of anatomy produced using traditional techniques of thermoforming.
You see we are dealing with the different problems of tactile graphics, design, manufacturing and application of them.
The following projects are actually underway. The first one is the application of an audio-tactile dialogue system for the blind which we developed two years ago. The second project is how can we support by means of special software the design of audio- tactile pictures. The third one is the production of tactile graphics using modified plotters. The last project is our so- called "hot" pen which is a pen working on swell paper. It uses the ink already available on the surface of the swell paper.
In my short talk I would like to speak about some experiences gained in the project of our audio-tactile dialogue system which is built up at the wall on my left hand side. But because the computer was only available at 8 o'clock in the morning we couldn't finish the installation by now. I hope we will fulfill this by afternoon.
I will speak about two things, firstly, some remarks about types of graphics for the blind, secondly, some remarks about audio- tactile dialogue systems.
1. Types of Graphics for the Blind
The central question is, as we all know: Are the texts, pictures, graphics and diagrams of the sighted people accessible to the blind? And we all know: The answer is "more or less yes" in the case of textual representations of information. And the other answer is "more or less no" in the case of two dimensional and relief-type representations.
Textual strings can be represented in braille or transmitted via the audio channel using speech synthesizers. Hardware and software are available to do this in sufficient speed and quality.
But, the first problems arise when text in papers contain nonlinear representations. Examples of that are two-dimensional mathematical notations, chemical formulas, and so on. Sighted people use this kind of representation because it possesses the advantage of a clear arrangement of the essential components of the formula. Why transform it into a linear braille form when its clarity is lost because of that? This is one idea of DotsPlus developed by John Gardner for scientific documents [Gardner 1993].
In such formulas, only a few very simple graphic elements arise, such as horizontal lines, symbols for sum, integration, and so on, which have a very close relationship to the alphabets of formal grammars [Loetzsch 1982]. But, the bigger the portion of graphic elements is in the picture, the greater are the problems to be solved when preparing the picture for the blind.
You can find numerous graphic representations, especially at school, at work, and during your studies. Beside the various graphics from basic disciplines, for example mathematics, physics, chemistry, biology, and so on, they are also used for building civil engineering drawings, arrow diagrams, structured programs, flow charts and so on.
Beside these representations related to education and job, you can also find a lot of interesting pictures, graphics, and diagrams, for instance, city maps; network plans of buses, trains, railways, and so on; maps for tourists; weather charts; surveys of parks and gardens; ground plans; surveys of museums; survey plans of zoological and botanical gardens; plans for exhibitions; and so on.
Of course, it would be very pleasant if one could get the tactile picture for the blind in a one-to-one manner from the original picture. But usually it is impossible to do so because of the complexity of the picture. An impressive example for such a graphic is a city map.
Therefore, in many cases, it is a long way from the prototype of a graphic to its equivalent for blind people [Burger 1993, Hinton 1993]. And, although this process can be supported by the computer, there remains a lot of manual work for which you have got have the right feel. Having not too many objects in the graphic means, for instance, you have to concentrate on the essential content. You have to leave out the unnecessary details. You have to split the graphic into several. You have to enlarge it and so on.
One piece of text gives an explanation to a certain object of the graphic. As the braille text must have minimum size, it is very often difficult to place it in the right position to the corresponding object. In such cases - you can replace the full text by a braille label and integrate all labels into an appendix with full text. - Or you can place the text in the area surrounding the picture and draw an arrow to the object (but this arrow is an additional tactile object, that must be recognized including at which graphic object the arrowhead is pointing). - Or, you can remove the text completely from the picture and write a guide of the picture which can be printed in braille or be recorded on tape. - Or, you can leave out the text assuming that a partner will explain the graphic object to you when you ask him.
Therefore, we can distinguish between the following four types of tactile graphics: 1. full text graphics, possibly with pointers; 2. labeled graphics with an additional legend, possibly with pointers; 3. guided pure graphics with an additional guide on braille or on tape; 4. pure graphics, then you need a dialogue with a person or a computer.
Undoubtedly the pure graphics possess an extraordinary importance when they are combined with a computer as a dialogue partner as we already have heard in the first lecture today.
A pure graphic contains only the graphic objects and no additional braille or additional arrows to be touched. Therefore, in that graphic you can place more graphic objects until the threshold of recognizability is exceeded. Of course, the computer partner is expected to guarantee the independent navigation through the graphic by the blind. Well, this leads us to the second point. I will talk now about audio-tactile dialogue systems.
He explores a tactile graphic in the usual manner using his fingers. If he wishes to get information about a certain object in the graphic, he requests it. Then he listens to the information given on the audio channel.
This leads us to the general scheme of an audio-tactile dialogue system: 1. There must be a place where the tactile graphic can be touched. 2. When putting the graphic there it must be identified by the system. 3. It must be possible to feel the graphic in the usual manner and to point to a chosen object. 4. The system must be able to identify the pointing position and the chosen object. 5. The system has to react by generating information and sending it through the audio channel. For that a speech synthesizer can be used.
The best solution would be to have a graphic display connected to the computer displaying the contents of the screen in tactile manner immediately. Unfortunately, until now, it was not possible to find a safe and cheap solution of such a graphic display for the blind. First steps in the right direction have been done by Schweikhart, Fricke [We'll hear something about his work today] Shinohara, Schultz and others. Therefore, at present, only one solution seems to be reasonable, the efficient design and production of tactile graphics and the preparation of them for the dialogue between the blind users and the computer.
In our work group, our first audio-tactile dialogue system which has been called Audio-Touch was developed in Dresden and finished in March '92. Since then, a lot of graphics have been treated for Audio-Touch from different fields of application.
Some years earlier Don Parkes from Australia developed his Nomad board which is now also manufactured by the American Printing House for the Blind [Parkes 1988, 1991; Duennes 1993]. Nomad and Audio-Touch, set the same goal but achieve it by means of different technical solutions.
At the wall on my left hand side, you can see the components of our Audio-Touch system: - a commercial graphical board to put the graphics on (It's simply a digitizer.) - a commercial speech synthesizer device that transforms the text of the explanation into an artificial language and puts it out over a loud speaker - and a control unit in the background which is responsible for the cooperation of all components. It's memory contains the Audio-Touch system itself.
Our Dresden Audio-Touch system can be configured for different digitizers and Touch Windows and for different text-to-speech synthesizers as well as for SoundBlaster cards with software synthesizers. It is working also on the Nomad board. The authors prefer to use a sound card with talking blaster because of the speech quality and the integration of sound.
Each common personal computer can be used for Audio-Touch. That means Audio-Touch is composed of standard devices. The only one which is special is the software.
You can use on the touch board, of course, different tactile materials: swell paper, reliefs of foil, paper embossed by means of a braille printer, and so on.
But, befor doing this you have first to design the audio-tactile pictures by means of the computer. For this we use standard software such as Designer, CorelDraw, and others or we apply a simple integrated module which is called Audio-Sketch. The implementation for Audio-Touch is then done automatically. I hope I can demonstrate the system this afternoon because we were not able to finish the installation procedure by 9:00.
The Dresden tactile plotter is a standard plotter, which has been equipped with a tank for special liquids. These liquids have been optimized in the laboratory, but not by us. And if you ask me which type of liquid it is I can't, unfortunately, give an answer, because I do not know the components of this liquid myself. After plotting the drawing on normal paper, the drawing must be stabilized by heating, because we want to make it swell.
The Dresden hot pen, as I mentioned already, is used on microcapsule paper. The simple idea is to take the "ink", already available on the surface of the microcapsule paper, and to make this ink swelling due to its heat effect. Stable contours arise which can be touched immediately after their formation. That means that practically during the process of drawing, it is possible to exchange ideas about the drawing between the sighted and blind people.
Thank you for listening.
A: It's done manually in the traditional manner. First building a matrix manually and then applying this procedure of thermoforming by means of heat and low pressure.
Q: Peter Korn. Berkeley Systems. I have a feeling I might not be the only one in the audience who would really love see a compendium overview or description of all the various technologies, their costs, their production times, etc. That I think would be really, really great if somebody here had that or if through the process of this presentation, we could put that all together. A list of every technology, who makes it, how long it takes, what it costs today, and any future comments from that company about where they're headed and what we might expect from them. I don't know who would be the person to do that, but maybe I could if everyone wants to just come to me I will just produce such a list, So, here I am. Come to me with your technology. I will make such a list.
A: I still use some pictures which have been made two years ago, and I have used it in many demonstrations and I could not find any degradations. It's very stable.
Q: Greg Vanderheiden, Trace Center. You had mentioned that your hot pen approach could be touched immediately as it was being drawn. How long does it take to draw -- to finish -- in other words how fast can the pen move across the paper, across the swell paper.
A: Yes. Very slowly. You can try it out. You can see it. You have to draw really very slowly. We made some experiments using this hot pen in a plotter and we had to take the lowest speed.
Q: So, roughly how long would it take -- what's an average drawing look like?
A: Oh, it's difficult to say. I have a demonstration paper with me and I can show it later. I know that we need about ten minutes for drawing it using the plotter. But, of course, it's a pen intended for manual use, not for plotting.
Q: And when you're doing with the liquid ink I would presume that would go much faster. How long would it take to generate one? I guess I'm starting to build the table over here. How long with the liquid ink? How fast does that run?
A: Because we didn't finish this project until now, it's hard to say with which speed we will operate the plotter. I expect a medium speed. Not too slow, not too fast.
Q: And after it's done, you stick it into an oven you said to cure it and cause it to swell? And what about--
A: It's not a very good technology and not a very good technique because you have two steps. The first one is drawing and the second one is swelling. I do not like this. Therefore it would be better to combine both in one device, but I do not like this additional mechanical work.
Q: David Wohers, Northeast Missouri State University. Could you tell me how hard you have to press down with the hot pen on the swell paper?
A: We didn't measure. We didn't measure the pressure, but of course, you must have some practice and a certain feeling of the right pressure. Please try it out.
Q: Did the plotter have to have anything added to press down on top of the pen or could you just mount the pen in the plotter and let it go.
A: No. No additional tool is necessary to do this.
Q: So really it's not very much pressure at all.
A: Ja.
Q: This is George Kerscher. RFB. Just a kind of a whacko idea, but the ink that's used in conjunction with the swell paper -- would it be feasible to use that on an ink jet, spray it, and then cure it, to speed up that process? Has that been looked at at all?
A: Firstly I will remark that we do not use the liquid in combination with swell paper. We use the ink on normal paper. And, second remark, of course, it would be nice to have the printer, the ink jet printer using this liquid but our group is not able to run this project because we are a very small group. But we are looking for people which are experts in printing devices and with them I think we could possibly build such a printer, possibly.
Q: Joerg Fricke. Fern Universitaet Hagen. Just a few minutes ago, I had the idea to use high energy laser instead of this hot pen. Did you think about this? It would be very fast I guess.
A: Yes. For a plotter I think this is surely the right solution, but not for manual use I think. (laughter) Any other questions?
Q: Greg Vanderheiden. Just an echo off that. Is anybody aware of anybody who's tried to activate swell paper using lasers? That might be an interesting thing to explore.
Q: My name's Tim Noonen from Rural Blind Society in Sydney. We're using swell paper and we've done some of it using computer production, but historically more using hand sketching and so on. What I'm amazed about with this swell paper is the smoothness of the background. Our swell paper always seems to have matte finish even on the areas that are supposed to be blank, and I'm wondering, is it our technique or the Japanese paper or the German paper that we used to use, or can anyone enlighten me on that.
A: We use a Japanese paper, but we are not very happy with it because, first of all, one sheet of paper costs about two Marks. This is very expensive and, therefore, we need another solution. I hope that we will have paper which is much cheaper as soon as possible. The surface of the swell paper after swelling is different. Perhaps it depends on the paper. Perhaps it depends on the copying process. If you can copy very fast, if there is certain copying device which copies so fast that no swelling begins, then you will have another surface. Possibly this is the answer, but I'm not sure whether you are content with it.
Q: Can you explain to us the copier used and the technology you use? Could you describe to us the technique that you use because that is quite exceptional compared with what we're used to having for swell paper. So, if you could just tell us exactly how you created it maybe we could find some clues.
A: Usually we use a Canon 5060. That's your standard device. A Canon 5060 for it. And it's fast enough for the purpose we think. But I think there might be also other copiers for that purpose.
Q: Then to swell it after you've printed it?
A: Yes. We swell after. After copying. Ah, we had a special swelling device. Excuse me, perhaps I didn't understand your question right. Yes. For swelling we use a so-called fuser. This is the well known rotating drum device. During one round the papers is swelled. It's not very convenient because it takes a lot of time for producing say 100 sheets of paper.
Q: Did you use a Ricoh fuser. Ricoh? And I purchased this in Europe.
A: We bought it some months ago and it was not a new device.
Q: My name is Paula Holmberg and I'm from San Bernadino. In the past, the encapsulated paper, the swell paper I've seen only raises to one even level. Yours appears to have raised to multiple levels depending upon how dark and how wide an area is black. Is that a special characteristic of your paper?
A: It does not depend on the black level. It depends on the texture. So in a small intervall you can produce different heights. We do not use any grey levels.
Q: Hi, my name is Mary Helmsom. I'm an independent cartographer and I do tactile graphics by contract, based on a contract basis. In a book called, Tactile Graphics by Polly Edmund, it's been printed through American Foundation for the Blind. I'm sure you're all familiar with it. She has a section in there on stereo-copying materials and with the rotating drum machine, the reproduction machine, you can get various heights with various, multiple passes through this machine. If you draw on the paper with your base drawing, put it through first, and as you put it through the machine once or twice, it will get a little bit higher, so you can do it that way too.
A: Okay. Thank you. My talk on paper? Yes, I have some copies of the paper with me. If you are interested I can give it to you.