New Graphic Displays for the Blind

An anonymous reader writes "Researchers from Spain have invented a new mechanism for graphical tactile displays for the blind. The displays use metallic films featuring various shape memory alloys which are produced layer by layer on silicon wafers using thin film technology. Display pixels are generated when the metallic film adjusts its curvature partially, similarly to bimetal snap plates for temperature switches. The movement of the films is then transferred to the touch panel via plastic pins und thus can be detected by the user."

Comments

Could this somehow work with colors/images?

[xmas2003]

Is there any way this could somehow do color/images? I read the story (yea, even though this is /.) and I think the writeup could have been more clear that I think (?) all this is doing is rendering text into Braile (the title of the story made me think it did more) ... but what about colors/images? I don't know what people blind since birth can "visualize", but for those folks who have had vision, could this be used as some approach to see colors?

I.e. I realize that something that is fairly "color deep" like my christmas webcam is probably undoable, but what about simple stuff like a red rose?

Re:Could this somehow work with colors/images?

[Justice8096]

I'm not sure that the impact of color could be interpreted easily with texture. Consider what it would be like if you tried to explain music to the deaf - The lyrics (if they exist) can be translated, the beat pattern and rhythm can be translated, but translating major, minor, augmented and diminished 7ths alone would fall short. Mathematically I can talk about ratios of 12th roots of 2, and poetically I can talk about "wholeness" and "loneliness" - but adding them all up would take a genius, if the person didn't have the experience of hearing (or, to put it another way, some of the interpretation is hardware-based).
Now, I can try to translate color into music, but that fails due to its dimensional nature - imagine trying to explain a theme with variations, where each different variation is played by a different player with spatial seperation - there aren't good enough words, and for the tone-deaf, it would still be useless.

Re:Could this somehow work with colors/images?

[zakezuke]

Is there any way this could somehow do color/images?

Color is a function of the human eye receptors. A person who was blind at birth would have no clue what blue, yellow, or red is. You could say the sun is yellow but that would have about as much meaning as the sun is *kikjij*. How to would one represent *kikjij* as a texture?

Tactile senses can only really tell you shape, texture (firmness/feel), and temperature. Color can not be directly translated into feel. A first step in translating the images into textures would be to work in monochrome needlepoint. You could translate your x-mas webcam for example into a large rectangle with a triangle on top decorated with balls of different materials such as metal, plastic, and wood. It would be a subjective effort but one which could at least communicate what is seen. All of this seem presently beyond the scope of these 10x10 modules.

Re:Could this somehow work with colors/images?

[DocSavage64109]

Instead of color, you could display a grayscale image by varying the depth of each pin depending on that pixel's brightness.

Maybe if you could let the blind user alternate the display from a "grayscale" channel to a red, blue and then green channel, the user could then try to visualize the scene in his head. I suppose a lot of this has to do with the resolution of the device, and how many varying levels of depth it can reproduce...

Re:Could this somehow work with colors/images?

[deathcloset]

more than color this could allow a blind person virtual 360 degree vision.

or anyone for that matter.

If you had a camera mounted on your back it could translate the image it captures into a type of virtual mosiac. It could do this via a body-wide graphical tactile display of high resolution.

then interpret/transfer that image to a "shirt" of this material - or a whole skinsuit.

So this way you could actually "feel" motion behind you. Perhaps you would even get very good at it too. It is not a far stretch to imagine that a person who has never seen could navigate in a chaotic environment just as well as someone with sight.

with the interpretation of the camera's image into "pixels" on your skin what's to keep the camera from zooming in if you wish it to? Imagine being able to feel a bird in a tree several hundred meters away.

A good place for graphical tactile response would be, of course, the hands but the lips are very touch sensitive.

And to that color point. If you were to transpose the camera's pixels into actual direct stimulation of the nerves in your skin then what is to keep you from having different sensations than pressure ?

Why not stimulate a stretching sensation? or I can imagine things that would best be represented with a soft sensation.....you know, puppies and kittens ;); Yes, this kind of research is great. It's immediate benefit is of course for the less fortunate, but it's real benifit is for us all. That's the best kind of science.

Why can't they just use the CLI?

[koreaman]

Why can't they just use the CLI? The only good reason not to is for multimedia, which obviously a blind person wouldn't care about, and multiple virtual terminals, which nowadays you can just do with Ctrl+F1, Ctrl+F2, etc. Why not use that instead of this presumably horribly expensive item?

Re:Why can't they just use the CLI?

[zakezuke]

Why can't they just use the CLI? The only good reason not to is for multimedia, which obviously a blind person wouldn't care about, and multiple virtual terminals, which nowadays you can just do with Ctrl+F1, Ctrl+F2, etc. Why not use that instead of this presumably horribly expensive item?

1) Brail terminals that i've seen only offer one line of text. They are already horribly expensive items. The diffrence would be using *nix mail vs mutt. Anything with cursor control doesn't work well with this or even phonics. It's damn helpful in lynx to have at least a 80*25 screen display.

2) While CLI would be cool for most things, this device would translate much of regular computer's display into a textured font making more existing applications useable. Something like XMMS or Winamp could conceivably be used on a tactile display.

Re:Why can't they just use the CLI?

[pere]

This is for use with CLI. As are todays braille displays.

Lots of fancy technology have tried to use tactile feedback for something useful for blind, but they have failed. The only useful tactile devise is really a braille display, and it displays single characters usually by moving 2X4 pieco-cells/pins up and down.

If you want to display a graphical interface, you really just map it down to one line of text (the line were your cursor is) - basically a CLI, and display it on an 80 character braille display.

Most other info (position, color, font-size etc) you either just discard, or display as sound.

Pin thingy

[l810c]

This kind of reminds me of a toy I've seen at Spencers and other novelty stores.

It's a bunch of pins going through a board that you can press your hand or face or whatever against and make an impression.(I did some googling for this, but the terms I could think of were too general)

If you could put a servo on each of those pins, it seems like you could pretty easily achieve the same result.

Re:Pin thingy

[ajlitt]

What's novel about this approach is that it is mechanically simpler and potentially cheaper to manufacture than current Braille readouts (which, as far as I can tell use many small solenoids). Also, since the display elements are bistable (that is, requires power only to switch up/down states, but not to maintain them) power consumption is minimal and portable PDA-like devices would become smaller and more practical.

Re:Pin thingy

[daltonlp]

Current braille displays use piezoelectric pins, not solenoids (pagers also use piezoelectric wafers to vibrate).

Piezoelectric pins are low-power, but also brittle. You can't get very good resolution because if you make them too tiny, they break. They also tend to have short lifetimes.

Solenoids are even bulkier, and draw more power, making them even less practical than piezo pins. Solenoids are good when you need a lot of punching power (which is why they're used for dot-matrix printer heads). They're overkill when you just want something a blind person can sense.

In 1999, two other Electrical Engineering students and I built something like this for our student design project. It used (guess what) shape memory alloys. It connected directly to the VGA output of a PC and averaged the color inputs to judge whether pixels should be raised or lowered.

It was only 8x8 pixels (just a prototype), but it was pretty awesome to move the mouse around and see those pins "do the wave".

The only drawback was the amount of heat it generated. Shape memory alloys change shape *because* of temperature difference (the change in temperature is not a side effect).

Even with a bunch of CPU fans cooling it, we were afraid our device would melt if we left it on for more than a few minutes.

It appears these folks have solved the heat/power problem with a design that requires power only when changing state. Nice work!

Interesting stuff

[mistersooreams]

You might wonder why these devices need to be so complex when Braille is just a series of dots. The thing is that Braille is a lot more complex than people think. (I think this is interesting but apologies if it's a little off-topic.)

Type I Braille is basically a 1:1 mapping of letter onto 2x3 arrays of dots. It's not much more than a font, but this is what people tend to think of as Braille.

Type II Braille uses a lot of abbreviations, and is rather more complex to read. For example, certain punctuation marks coincide with word abbrevations, and only the context serves to differentiate the too.

Type III Braille is still more complex and is almost like a whole other language. I don't know much more about it than that, but anyone who does can add to this.

So you see that the increasing complexity of these devices actually makes life a lot easier for blind computer users. I wonder how many blind people read Slashdot?

Re:Interesting stuff

[Anonymous Coward]

Well, standard braille uses a 6-dot matrix which will represent 63 different symbols.
So, in grade I braille there is a one-to-one mapping for letters and punctuation
but for numbers the letters a thru i represent 1 to 9, j represents 0; but numbers are preceedded by a reserved "number sign" pattern to
disambiguate. In grade II braille, things are a lot more compled. You have "contractions" where a single cell is used to represent frequently-used letter combinations like (sh) or (ch).
These contractions can also be position specific, e.g. one particular symbol represents "dis" at the beginning of a word, "dd" in the middle of a word, and "full-stop" at the end of a word.
You have single or multi-letter abbreviations, e.g. "p" on its own represents "people", "nec" represents "necessary". The rules go on and on.
However, this complexity has nothing to do with the hardware. It's handled in by software whenever grade I or grade II output is needed.
Blind hackers like myself use 8-dot computer braille which generally has a one-to-one mapping of ASCII to braille symbols.
(and I'm only posting as AC because big fat lazy Cowboy Neal is too big fat and lazy to get up off his big fat lazy arse and answer my e-mails asking for help setting up a Slashdot account. I can't do it myself because I can't read the captcha. (maybe if I had one of these new hi-res thingies capable of rendering the shape of it ... what's this goatse thing about anyway?

is this the best todays technology can do?

[r4d1x]

Why not just hook them up to the BrainPort? A step closer to helping them "see" again......

Audible Graphic Display?

[devnullkac]

bimetal snap plates

I think it would be interesting if the pixels actually made an audible snap when they change. I don't have any visual disabilities, but it would seem that Braille offers no equivalent to the peripheral vision sighted people use to take full advantage of a large graphical display. Such snap sounds (if done subtly enough) could be a small step in that direction. A "multimedia experience" of sorts for the visually impaired.

Then again, Braille terminals may already have this: in the movie Sneakers the terminal used by Whistler was making sounds as it was updated, but that may have been artistic license by the director.

Re:Can not represent pictures as per see with bump

[tsa]

In this whole discussion it is forgotten that most blind people are not blind from birth. They would certainly benefit from the use of 'colours' in this way.

Re:Can not represent pictures as per see with bump

[greenhide]

If you brush your hand over your keyboard or mousepad and close your eyes can you not construct an image?

Do you really think such an ability is only the relm of the sighted?

Certainly they with no sight will have a different image, but will it be any less nuanced?

You're using the terminology of the seeing. I'm not blind; I have no idea what goes on in their minds. But, I imagine if they've never seen with their eyes, then they wouldn't have a "picture" of the keyboard, they would have a mental map. And it would be different than how we comprehend our own mental "images" of things.

Let's think about things that you never actually see. For instance, what is your image of what a headache looks like? Assuming you've never seen a picture of it, what does a stomach look like? After all, you have one in your body, and you've certainly felt it at times in your life.

Here's a slightly more risque one: if you're a guy, ever been inside a lady? Would you be able to draw what the inside of *that* lady looked like just by how it felt? I doubt it. But I'm sure you have a mental map of it.

Or more banal: ever stuck a q-tip in your ear canal? Would you be able to draw what it looks like? Possibly not, even though you've traced its contours with the q-tip. And yes, I know no one should ever stick a q-tip in there, but everyone does it.