Catching up to the Future: Charting Your Own Courses

Editor: This TDI workshop was presented by Dr. Gregg Vanderheiden of the Trace Center at the University of Wisconsin in Madison. He discusses two "futuristic" ideas about how a person might interface with various devices.

~~~~~~~~~~~~~~

It's sometimes hard to predict how fast a certain new technology will spread. An example is the cell phone. When they first came out, the law exempted them from the requirement to be hearing aid compatible (HAC), because they were considered to be a specialty item. At the time, no one envisioned them becoming a mainstream product, so HAC was not considered to be a requirement.

Another example is the mushrooming of the Internet. Because it was such new technology, no one really understood what it was or how it should be regulated. So Congress left it pretty much alone. That's a problem now.

An example of technology that hasn't developed as fast as we expected is speech recognition. For thirty years now we've been hearing that reliable speech recognition is only three years away. Now it's useful, but it's still not as good as we would like.

IBM is working on "superhuman speech recognition", the goal of which is to be able to outperform a human. I wonder if they are considering things like "deaf speech" or people with speech disabilities. Or will this new technology work for everyone except those with disabilities?

I once visited a research organization that was developing a program to check English grammar. Someone wrote out examples of ASL grammar [Ed: which is very different from English grammar] and fed it to the program. The program reported that there were NO errors. The reason is that the developers had programmed it to look for grammar mistakes that English users commonly make, and the examples didn't contain any of those. But they were a long ways from grammatical English!

Siemen's has developed a Virtual Keyboard, which projects a picture of a keyboard. A user types on the projected image and the Virtual Keyboard is able to capture the intended keystrokes.

Several companies have developed virtual displays. The user wears a pair of "glasses" and sees a display floating in front of him. The display could be used for captioning.

Another exciting new technology is direct brain interfaces, in which electrodes are inserted directly into the brain. The user is able to control a joystick just by thinking about it. This is technology that paralyzed people can use today!

We currently have a tiny pill that contains a camera. Potential applications include an alternative to a traditional colonoscopy. We will soon be able to inject a device into a vein and have it go to the brain, where it will direct brain connections.

Today for $1000 you can buy a computer that has the processing capability of a dragonfly; by 2023 $1000 will buy the processing power of a human brain; by 2044, $1000 will buy the computing power of everyone on the planet.

Universal Remote Console

The potential development of the Universal Remote Console (URC) shows how being involved with a project from the start allows a person to make huge fundamental differences.

Suppose anyone could control any device from anywhere using whatever device is handy. The International Committee on Information Technology Standards (ICITS) is working on this, and they are looking at using standard technology to do it.

Suppose every device provided a virtual interface into which a user could plug whatever device she liked. That concept provides a lot of capabilities. It could easily include things like language translators, for example.

The key to this is to be sure that it's universally accessible; if the only interface is speech and you happen to use ASL, you are completely excluded from the technology.

Personal Services on Demand

This concept provides remote services from any location. For example, a person could send an audio stream somewhere, have it converted to text or ASL and sent back. This is currently being done with a human in the loop, but imagine if it could be an automated service. The current problem is converting speech to text. Once that is overcome, the service will be available.

VoIP

When you make a phone call now, there is a continuous wire from your phone to the other phone. VoIP (Voice over IP), which uses the Internet to transmit phone calls, is just becoming mainstream. The attraction of VoIP is that we don't need separate phone lines throughout the country, because the phone conversation is just another type of data. Multiple people can conduct different conversations at the same time using the same infrastructure. This flexibility makes it easy to bring someone in to assist if you have problems understanding (e.g., an "on-demand" relay operator).

Problems?

Here are some of the potential problems to be aware of:

- TTYs may or may not work with new technologies. It may be that one works with the new stuff and another doesn't.

- New technologies may not work with each other.

- Public phones may not be accessible, or the type of accessibility it provides may not be what your device uses.

- Do any VoIP phones have to be accessible? Some people are working hard to make VoIP accessible.

A person may be charged more for a voice plus text call than for a voice call.

There are two opposing dangers concerning TTYs. One is that by trying to maintain TTY compatibility, we stunt the opportunities to use new technologies. The opposing danger is that by trying to maximize opportunities with new technologies, we lose compatibility with older TTYs.

Doing it right means that we create standards and technologies that allow anyone using any technology to communicate with others using any other technology, and the cost will be the same for all.

Projects We're Working On

SIP-TEA is a VoIP Client that simultaneous text, speech, and video. It's a free program that runs on a standard PC.

T-Server provides interconnection between different technologies, including standard telephones, VoIP, wireless, etc.

Our goal is to allow as many devices as possible to interact without requiring transcoding device. It should all be "built in" to the system.

T140 is a protocol that may replace Baudot [Ed: The standard TTY protocol]. We need to replace it soon, because it really limits interoperability.

Standards are KEY; if we are to have universal connectivity, we must have standards and devices must adhere to them. We must get accessibility incorporated into the standards.

Standards and regulatory bodies are having a hard time understanding the problems. We need to be involved in the process to make sure it works. We need deaf people to tell them what deaf people need.

We all need to stay informed and involved, because many quick solutions are traps. The people making the decisions need to be exposed to many different ideas and perspectives.

Q: Are you working to be sure that location information travels with the packets?
A: With IP, you don't know where it's coming from. 911 calls using VoIP give no location info.

Q: Deaf people seem to be more willing to keep up with technology than hearing people. Do you agree and how does that affect things?
A: Deaf people may be more adventuresome because they have more to gain. Adding video to a phone is not very useful to a hearing person, but vital to a deaf person. The best way to get accessibility in is to have the feature desired by both deaf and hearing.