Cost 219ter

Proceedings of

Extending Horizons

16th January 2007

Conference organised by COST 219ter

Accessibility to Next Generation Networks

DR JOHN GILL: Good afternoon, can we please take our seats. I hope you feel suitably refreshed, having had lunch.

We now have quite a busy agenda for this afternoon. Our first speaker is Professor Gregg Vanderheiden, who is a Director of the Trace Research and Development Center in Madison, USA. Gregg has been around for probably longer than I have, myself and Alan Newell. We welcome you, Gregg, and look forward to your presentation.
 
PROF GREGG VANDERHEIDEN: Thank you. The value of relying on technology here! Well, in a few minutes we will have the slides up here. I will be talking today about the process of trying to develop guidelines and/or regulations, if you will, dealing with accessibility. In different countries they are called different things and take different forms. The interesting part about trying to do this, however, is to try to write guidelines that are going to be useful today and still be useful tomorrow. We can either write them that they will be useful tomorrow, but they won’t work today, or we can write them for today, in which case they become obsolete rather quickly.

This work is based on the work that we have done with over 40 different standards groups in Europe, the United States and around the world. Also, we have been working on accessibility guidelines since the early 1990s, when we put our guidelines for consumer products, then we worked with ITF, IBM and Microsoft in developing the guidelines for software and hardware systems.

We did the first web accessibility guidelines WWW2, and we have been working with COST also throughout this whole process.

A number of issues have come up. I want to talk about the issue and then we will talk about some strategies that we might use for trying to address these issues in a way. Now, the first one is that every time we sit down and work on these together between consumers and industry, industry always cites two constraints. They’re both very real. The first one is that guidelines must be general. If you make them specific, then industry can’t innovate. The second one is that the guidelines must be specific, because if you make them general, there’s no way to figure out if you’ve actually met them. You can make general ones but you cannot say, “Here’s the general guideline and you must do it.” Because they say, “Well, how will I know? You will get mad at me if I don’t do it, but how do I know I’ve not done it if not specific?” You cannot require that I do something if you can’t tell me when I’m done.

These are both very, very real. We’ve worked with over 50 companies and we work on the design of their products. It becomes absolutely clear that both of them are very real. How do we go about trying to address this? I would like to try to suggest that we do a combination of testable standards with sufficient techniques. I mean that we start off with general but testable guidelines. They have to be testable and decide that they have been met, but you don’t say exactly how they have been met on each particular platform.

You couple with this sufficient techniques so that you would say, for example, that you must do the following and the following are examples of sufficient techniques.

In the web content guidelines, we might say, for example, that you must provide captions along with the audio description. Then we would say, if you are using this particular technology, this would be a sufficient technique. Or for telephony, you must say that you must have a mechanism for real-time text on any voice phone calls. Then you would say on the PSTN, it would be sufficient if you supported V.18 or whatever it is. Or, IP network, it would be sufficient if you did the following, etcetera.

The thing about sufficient techniques is that you don’t have to necessarily follow that particular sufficient technique. You could say this is sufficient, but you also allow industry to innovate and have some other technique that they also could implement that would still meet that testable guideline.

So the testable guidelines tend to look a little technical. But with the sufficient techniques, you can make it very down to earth.

If you look at the web content accessibility guidelines, they tend to say things like “the information must all be programmatically determinable by user agents”. You ask what does that mean. If you use HTML, this is automatic. If you are going to write Flash, you need to expose it to SMAA. If you are using flash, if you are not, it doesn’t really matter. But it is very concrete within each of the realms that you are talking about.

It allows then industry to come back and say, well, instead of, for example, doing closed captions, what if I made open captions, or what if I implement the technology so that it was successful in the following fashion. Then they are home free. They did what you suggested and everybody agrees. If they do something else, then they have to justify the fact that that is in fact sufficient.

So you have an easy path but you can still innovate. If you have to innovate, you have to prove that that is sufficient, and, over time, your innovation could be accepted as a sufficient technique going forward. So it allows you to innovate overnight.

One of the problems we have is that companies will say, “I want to innovate, I want to announce on March 1st and not tell anybody before then that I’m going to have new technology. But I don’t want to not be able to sell anything for six months afterwards because this thing has not been approved by somebody.” With this approach you can. It is just that on March 1st you are going to have to have some documentation. You are going to have to have worked with AT vendors or something. You are going to have to have done your homework to demonstrate that this is successful.

It provides detailed techniques after what you can do without having to do detailed investigations yourself. You know what it is and you can just do it. It allows you to do other things as well and allows you to do sufficient things that change over time. This is critical.

Today, it may be that you need to do X. But in the future, as AT gets smarter, you may need to do less. Less may be sufficient in the future. So it allows the guidelines to stay the same as to what they are trying to achieve but allows them to evolve over time and move in the direction it needs to.

Warning, big warning. There is a dark side to the sufficient technique approach. That is, if you go through a very careful process of developing your guidelines, involving consumers and everything else, but you allow the sufficient techniques to be just casually established, you can actually have people establishing – In the United States, we had a famous instance where it was required that you had a vegetable at lunch. So they decided that the pickle and the ketchup on a hamburger were vegetables! That was sufficient. Suddenly, having a hamburger met all the dietary requirements. A pickle was decided to be sufficient vegetables.

That is the kind of thing that you have to be careful of. “Sufficient” has to go through the same kind of care and development as the rest and consumers need to be involved so that, in fact, it is sufficient.

Number two: making the guidelines stable over time.

Over the decades technologies, the barriers and solutions, change greatly. Barriers we used to have are not there any more. We are inventing new ones every day. But people and their disabilities don’t change very much, our abilities.

If we base the guidelines on function and we also talk about characteristics rather than trying to sort them into categories, we would be much better off. It is impossible to say, “These are the ones for phones, for PDAs and these are the ones for computers.” You cannot tell the difference between the three any more. Even, you cannot tell the difference between a telephone and a television set and a computer, as you have seen today. If you have not seen it today, then looking at the market place you will see it.

Also, we can identify the key underlying strategies that we use and build off those. We use what I call the “bow tie approach”. You have a large number of user needs at one end. The Trace Center developed user-needs guidelines extended by the Special Working Group. It has 100, 200 needs that people have in the area of interface. But it turns out that with a smaller set of strategies you can address those needs.

For example, if you have voice output, it not only helps people who are blind but also people who have cognitive disabilities. We have the basic strategy functional requirements. You allow industry to then implement a variety of solutions based off of those.

You need to provide a way for someone who is blind to be able to operate the device with controls they can find tactically, discernable controls, that allow them to access all the functions using tactically discernible controls and voice feedback. There is a whole bunch of different ways of doing that.

You say if you do that, it is a sufficient way for allowing, or that can be a requirement for access, It turns out that for some individuals that combination is the only thing that is going to work.

The next thing that we need to pay attention to is that the whole field of mainstream information and communication technology and assistive technologies is changing. We’re finding that there is no clear definition for assistive technology any more. Screen-readers used to be assistive technology, but now Apple builds it into their operating system. It talks. It allows anybody to have information read and operate it without the screen. Is that assistive technology or part of the operating system?

More and more, we’re finding that these things are happening. People say it is still AT. You say why. They say otherwise my formulation for how it has AT compatibility doesn’t work, if it is in the operating system it is.

We are finding that AT and universal design are blending. We now have cellphones that you buy and you want to change the ring tones. You download them, right? They were not in the phone. You scroll down and push a button. It says “ring tones” and it downloads them off the web or network as you are going along.

You can add features to your phone without knowing that you have added to them. When you selected them from the menu, it loaded them into the phone and that was the first time that feature was in the phone. But, as far as you were concerned, they were always in the phone. We are finding that half our products are in the hands and half on the network. We are going to see more and more of that kind of thing.

What we need to be looking at is the fact that we are not going to be able to draw a line about what is AT and what is not AT. So defining our accessibility around AT is a problem.

The second thing is that AT is not always allowed. So you can have a computer and say, well, I made my programme accessible to AT. You find out that, in fact, it cannot be used by anybody who is blind, because, in fact, it is being used in the library and they won’t let you install any AT on it. So the fact that it is AT-compatible has nothing to do with access.

We have book-readers that will read any book, except if the author, the publisher, checks one little box in the data structure that says “don’t read this book aloud.” The book-reader can read it but it won’t. You have something that would be an accessible product, accessible book, except that with a check it is not allowed to read it to you.       

We also have AT that is so expensive that only some can afford it. One company wants to release a website and they’re releasing it with new technology. On the same day, they worked with an AT vendor. The AT vendor will release a screen-reader which will work with it. So nobody in the world will have a screen-reader that can access that content on the day it is released. But it will be accessible.

The AT also costs £1,000. That is out of the reach. So what are we going to do there?

We can add products that are closed, and we’ll talk about closed products as being any piece of hardware or software the user is prevented from installing, attaching or using AT due to physical, electrical or policy consideration.

We have a couple of interoperability standards working along, people working on APIs that would allow you to install software that works with other software and can read it. Right now, there are some de facto ones in different platforms. There are some universal and remote console standards that have been passed. They are ANSI standards, and ISO standards are up to F disk. That would allow you to substitute the interface on a product. So the product would be otherwise closed but you would be able to use a different set of buttons and displays.

This can really change the world. So the way to address this, we believe, is to define it in the following way.
      
A product must be accessible when it's evaluated with the AT that most users have when they encounters the product, and of course they are allowed to use.
      
So if you have built-in access and AT is not needed you pass. If AT compatibility is used to say this is accessible because it's compatible with AT then you must demonstrate that in fact people who come up to this will have the AT that it will exist your users do have it if your users don't have the AT you can't get it and call it accessible.
      
For productivity settings that is work stations etc, we really need to be looking at AT trying to build access in doesn't allow users to have the custom interfaces.
      
So this gives a kind of three layer model for accessibility use mainstream products wherever possible to make things accessible that's always the nicest everybody goes up and uses the same product. Can you build in compatibility with personal assistive technologies and make sure when they come up to the product it doesn't jam their hearing aid, or interfere with the use of their products, and thirdly compatibility with adaptive technologies we can talk about that.
      
Now, the last one, this is the one where it's kind of a downer and it's a challenge as well.
      
We have now reached the point where we have a choice of stifling innovation or leaving people behind. We have two choices right now, we can either let the old assistive technologies drag the market, and it turns out that there's a large part of the market that doesn't have the latest AT and the second latest AT and they have the third, fourth and fifth. We have more and more web technologies and other technologies that are only accessible using the latest technology. Many times it costs $1500 or more for that latest technology so we have to either say well we are going to use any new technologies but that's not acceptable. Or, we have to say that we are going to go ahead and use the new technologies and just forget about that other third of the disability population, if you don't have a job, and you don't have enough independent money to be able to afford $1500 which you very often have to buy again every two years because of the role of new technology - rollover in technology. You aren't going to be able to access the content using the new technologies. That's not very acceptable.
      
Industry needs to be able to move forward, industry needs to be able to use new technologies, we want to be able to use the new technologies. We also want users to not be left behind, but what we are finding is that rising water floats all boats, yes well, some float higher than others. What we have is, as technology goes ahead people with disabilities are rising but everybody else is rising faster. And, this gap is the problem, if you have a job, that gap can prevent you from being competitive on the job. There are always going to be some at some disadvantage and the comparative advantage is so much different that we are in fact having people who are losing jobs, and they are being moved out of all the critical paths because you can't have somebody in a path where they really can't keep up and not getting promotions and can't get into new positions and that's a real problem. We need to be trying to address it.

So we can either let the market run, just call new technologies accessible, if they work with the new stuff, and the fact they don't work with the third half of the people, that's just too bad. We let those people fall behind.

Or we only use the technologies that work with the old AT. You would not believe what doesn't work except with the old AT and that really would put the web and a lot of our new technologies back. S we need to find a third choice and find ways to provide access to ever-newer technologies for users with disabilities who are unemployed or underemployed and have few resources and just have to figure out how to do it.

Now building-in access is one way and then they pay the same thing for everything as everybody else. But we can't always do it that way. You can't make a product that somebody doesn't need a wheelchair for, and can't always build access in. We could fund updated AT to those who can't afford it, some countries do this and take the people who need the advanced technology, and provide it to them, and in fact they probably save money because by providing the latest AT to those individuals at the bottom who can't afford it allow the whole thing to shift forward, and they reduce the burden on all the way up. That's another one.

The last one I would like to throw out is the idea of providing virtual AT. That is that an individual who is blind for example could come up to any computer that has speakers, and somebody could help them log into a website called I don't know "virtual screen reader.com". From there, it would start talking, and then they could browse the internet from there and their screen reader would not be on the computer in front of them but on that server in the network and we would maintain these servers in the network, and they had the AT on them. And we would keep them updated so they work with the latest AT, but they won't work as well as AT installed on your own computers. Those people who have jobs, they would still use the AT, but it would allow individuals who are not able to use that to be able to have the virtual IT that would always be present, and present any time anybody walked up to any compute.

These are three challenges. We need to be looking at this. If we don't address this problem we are in fact going to leave a lot of people behind or dragging everything back, and neither one of these is acceptable choices. Let's dedicate ourselves to trying to figure out good solutions for the third choice. Thank you.
             
DR JOHN GILL: Thank you very much for a stimulating presentation Professor Vanderheiden. Our next speaker is Inmaculada Placencia-Porrero (which I find great difficulty in pronouncing and I tend to shorten it to Inma). We remember her from the days of TIDE which had a close association with COST and she has now moved over to DD employment.

Previous

Contents

Next