PhoneAbility
Location-based Accessibility Information among Communities of Interest
Mr. John Peifer
JOHN GILL: Our next speaker is John Peifer, who is Research Director of the Biomedical Interactive Technology Centre of the Georgia Institute of Technology in Atlanta USA. We are extremely grateful to him for flying over here today.
Since 2001 he has been the Co-Director of the Wireless Rehabiliation Engineering Research Centre, RERC, working to improve community access, independence and quality of life using mobile wireless technologies. Mr. Peifer.
MR JOHN PEIFER: Good morning everybody. I want to express my deep appreciation to Dr Gill and the sponsors of this seminar for all of their efforts in co-ordinating this event. It's very exciting to bring together all of these people to share ideas and to explore these new technologies. I also deeply appreciate Dr Gill for inviting me to participate in this event.
I am John Peifer and I am the Co-Director of the Wireless Rehabilitation Engineering Research Centre in Atlanta, Georgia, and this morning I am going to give you a brief overview of some of the work we have done on developing research prototypes of mobile wireless technologies for sharing location based accessibility information among communities of interest.
Professor Jackson earlier gave an excellent overview and introduction of some of the technologies that enable these applications, and I think he was very accurate in pointing out that one of the major limitations today is in the human computer interface. Making the information available and easy to access is a remaining challenge.
So today I am going to tell you some of the things we have tried and some of the lessons we have learned. A brief outline of our presentation this morning: I am going to give you a quick description of the wireless RERC and then go into some more detail on one of our wireless RERC projects, the mobile accessibility guide, which led to the work in sharing information among the communities of interest. Then I am going to talk about two other RERC projects that are related to this and how they complement the mobile accessibility work. Finally, I will conclude with some of the challenges that we face in location based applications.
Wireless RERC is short for the Rehabilitation Engineering Research Centre on Mobile Wireless Technologies for Persons with Disabilities. We are one of 22 different national RERCs sponsored by the National Institute on Disability and Rehabilitation Research (NIDRR) within the U.S. Department of Education.
The Wireless RERC was initiated by NIDRR because they saw the emergence of wireless technology as a way of life, and they were concerned that if people didn't have access to that technology then they would be prevented from participating fully in school, work, recreation, and community activities. There was a need to promote accessibility of these new technologies. At the same time there was a great potential in using wireless technology to improve independence and quality of life.
Therefore, our two parallel goals within the Wireless RERC are to promote accessibility of these new wireless technologies and to explore applications to promote independent living and community integration.
We have a collection of research and development projects. I will mention them briefly. Three main research areas include 1) an assessment of user needs, 2) evaluation of emerging technologies, and 3) policy initiatives to support universal access. In research project 1, we have established a consumer advisory network and developed a test model process for predicting usability of future products and services. Research project two is to evaluate some of the emerging wireless technologies and to work with industry representatives to determine how new technologies might impact accessibility; identifying barriers and opportunities. We are trying to encourage industry leaders to incorporate principles of universal design to provide universal access for their emerging products. We are all ageing into disability and industry will miss a major market segment if they only build products for 20-year-olds with nimble fingers. Our third research project is policy initiatives to support universal access.
We have five developmental projects to explore what you can do with wireless technologies. Project one is to explore multimodal interfaces and alternative interface standards. Project two, a wearable captioning device, provides wireless text captions for deaf and hard of hearing. Project three, mobile wireless communication access, combines wireless communication and Internet access with alternative and augmentative communication (AAC) devices. In addition, we have explored using some of the new location based technology to provide context for these communication devices. Project four, a mobile telerehabilitation project is exploring mobile rehabilitation and community re-entry applications. Project five, we are looking at wireless technology as cognitive prosthetics: how we can use mobile wireless devices to provide time management, wayfinding, prompting and cueing.
Now I would like to talk to you about the mobile accessibility guide project. The concept for sharing information grew out of this project. The idea was to apply mobile wireless technologies to distribute information about community accessibility where wireless networks could deliver information any time anywhere. With location we can add context to the information delivery and simplify the interface.
Accessibility information is already widely available on the Internet. It's valuable for avoiding problems and dangerous situations, and allows people to plan trips and explore in the community. With wireless technology, we can bring information to people while they are out and about and can increase confidence and spontaneity while travelling.
An important point is that, with wireless technologies, individuals can become mobile reporters. They can identify problems and solutions while they are out and about. We believe this is a powerful tool to update and sustain a current database with consumers forming the community of interest.
This figure below is a picture of an early prototype of the mobile accessibility guide (MAG).
This first version of the MAG consisted of an iPaq PDA that has a display with text and image capabilities as well as audio capabilities. The device also included wireless networking and position tracking. The mobile accessibility guide concept is to use the wireless network and location tracking capabilities to collect and distribute accessibility information. Position tracking on that device helps to determine when to collect information and what resources to present to the user.
The mobile wireless network interface could be a WiFi wireless Internet connection or it could be a cellular data connection. There are also new wireless network technologies emerging, such as WiMAX and ZigBee - and it’s reasonable to predict that we will have wireless Internet access everywhere very soon.
New position tracking technologies are also rapidly emerging. GPS is widely available outside, where you are clear of buildings, but GPS is improving, and new emergency 911 services in the United States for cell phones, RFID technologies and other location tracking devices are becoming available. We can expect these trends to continue, producing better location tracking capabilities, indoors and out, on many mobile devices.
There are different types of information that support location based services, and it is helpful to consider these resources asinto several layers of databases. As a foundation, start with the geographical information system (GIS) databases that describe streets, maps, buildings, parks, shops, and different kinds of landmarks.
There are already many examples of established GIS databases that are now being tailored to car navigation products, and they are rich sources of information that can be used for other location based services too. But many of these mapping resources represent static information that isn’t updated regularly.
The second layer of database organization consists of published accessibility guides which describe resources, ratings of businesses and services. Does a restaurant have a ramp? Is there a braille menu available? Those kinds of accessibility guides are available on the Internet and we can tap into those and bring them to a mobile device.
The third level I have highlighted would be consumer annotations. This is the community of interest, where people can share their own descriptions about community accessibility. They can provide updates on published ratings. Maybe a restaurant received a great review for accessibility but they have closed the ramp or shut down the elevator or the braille menus are out of date. The consumer can update this accessibility information. They can provide advice and warnings, and recommendations for things they like. They can provide descriptions, directions on how to get around a problem, or a comment about encounters with sales people or assistants.
Other presentations in this seminar have provided examples of emerging products using location based technologies. We are going to see more and more of these, so what we want to do is build on that emerging market of location based technologies, location based or location aware devices, and encourage these companies to think about accessibility and share that information with communities of interest.
Accessibility resources are in print and on the desktop. This is just a short list of some of the websites we have available in the United States for planning travel. Different states have information about accessibility. Different venues such as DisneyWorld provide information about how to get around the park and what kind of resources they make available. These resources are valuable for planning a trip, but we think it's also valuable while you are out and about to have that presented to you on your mobile device, and you can change things and find out about accessibility.
The Shepherd Centre in Atlanta, one of our partners in the Wireless RERC, have provided a very good Guide to Atlanta for People with Disabilities in which they publish expert reviews of different places and services in and around Atlanta for accessibility. What sort of resources are available and how accessible are they? The guide describes three accessibility levels: 1) do they meet the ADA requirements, 2) do they provide basic accessibility, and 3) no accessibility. Information is provided on places of interest such as hotels, restaurants, banks and places of worship; as well as commercial services such as equipment repair, home health, and transportation. Accessibility of these places and services is reviewed features such as parking, entrances, restrooms, and seating. While this is an excellent guide, it becomes outdated as restaurants close and conditions change. That's where the power of wireless technology comes in because it enables us to update this information any time and to deliver it anywhere. These ratings and authenticated reviews in the Guide to Atlanta become more valuable with consumers updating this information.
Mobility increases spontaneity and independence, so you can change plans while you are out and about. If you find your favorite accessible restaurant is closed, you can use the mobile accessibility guide to find another accessible restaurant that is nearby. While you are out and about, you can report problems that you encounter and provide advice about how you solved those problems. Sharing that information can be very powerful for others with similar needs and interests to you.
Location awareness can simplify the application interface by filtering the information. This is important. While a location aware device is more complex, that complexity could be used to simplify information retrieval applications on that device. If the device knows where it is, it can anticipate your needs if it has an accurate profile of your interests and needs. It can be designed to deliver only the information that's relevant to you at that particular time and that particular place.
The prototype that we developed for the MAG had several different modes and key features. We had an interactive mode, where the user could request information about places and resources that are near (such as shops, restaurants, or banks).. A second mode was an alert mode, similar to a customised weather or traffic report, where the device could be programmed to deliver information to you based on what is near you or to provide information about changes in an area that you often visit.
For example, somebody may have reported there was ice on the bridge that you often use. You could have designed into your personal profile that that is a route you use every day. If ice occurred on that bridge and another consumer reported it, it could be brought to you as an alert while you are at home, before you head out, or while you are traveling as you approach the bridge. Even if it's not part of your route, the device knows where you are and knows that somebody reported a problem of ice on the bridge, and that could be delivered to you as an alert, with a recommendation perhaps for how to avoid the ice. The report might include a recommended path to get around that problem. Again, what makes this work is that the consumers send updates to the database, to expand the content and keep the accessibility information current.
For another example, the mobile accessibility guide could show a picture of the 13th Street Christian Church steps with a text and audio note explaining where to find an entrance with a wheelchair accessible ramp. The way this information is presented would depend on user preferences. Some may prefer an audio presentation, and others may prefer text, pictures and audio.
In the alert mode, the PDA shows the message, "Alert! No wheelchair access to the diner until elevator repaired." In this example, someone has reported the broken elevator to the shared database of community accessibility. It may have been a diner that they regularly visit, but wheelchair access is impossible without the elevator. This kind of information can best be kept current if consumers are willing and able to share reports on community accessibility.
So, this is similar to a customised weather or traffic report, for your own route, your own neighbourhood, for your own interests. The key here is how do you develop intelligent filter algorithms to present the information only relevant to you, using date, time and location preferences?
Some of the things we have explored are how you define the search radius. How far do you want to look for accessible restaurants, your shopping interests, transportation methods, routes between home, work, church, recreation, and what types of abilities or impairments do you have.
A variety of factors can go into the personal profile for filtering relevant information. What sort of barriers to access are you concerned about? What is your level of health and endurance? If someone is using a wheelchair and they choose to use a manual wheelchair, the slope and the length of the ramp could become very important, so that would be something they would want to share with others, but that may not be relevant at all to other consumers. What types of assistive technology are you using, and what sort of device characteristics and presentation capabilities does your mobile device have? Does it have a braille display, only audio, or audio and graphics?
Consumers can become mobile reporters who submit annotations about community accessibility, and the process can be very simple. Press a button to record your message. Press another button, take a picture. Entering text is more of a challenge. On the mobile phones you can do that, but the multi-tap methods are difficult. Speech to text would be a way of entering text but location aware devices allow you to tag the time and location to that annotation, so simplifying the input is also a key to making this a success.
People aren't going to document problems or solutions if they have to spend 15 minutes constructing a report. How much time will they spend to share reports? Is it 15-20 seconds? We don't know the exact answer to the question, but it has to be very quick and easy for consumers to report problems and share advice.
Our initial prototype of the mobile accessibility guide was expensive and complicated. We used an iPaq PDA that had a slot for a Sprint data card that gave us a wireless Internet connection. The PDA had a Bluetooth connection to a GPS device with a separate component you could put on your shoulder, or carry in your pocket or your hand, but you had two different pieces of equipment to juggle. The PDA itself became rather fat because we added an expansion sleeve with a slot for a camera that could take pictures.
Then we had to write the software to integrate all of these functions and it became more complicated. Before using the device, it was necessary to load custom software, and then software conflicts led to rebooting.. While it was interesting from a research standpoint, it was not simple or reliable enough to release for average consumers. People would be unlikely to use such a complicated system that involved assembling multiple components and loading custom software.
New off-the-shelf mobile devices support most of the features that we established with the initial MAG. Camera phones are widely available to submit pictures. The wireless Internet now is built in - WiFi and cellular data connections are available and the wireless Internet access is becoming widespread - so that problem is going away. There are still problems with accuracy, but more off-the-shelf cell phones and mobile devices are providing location tracking capabilities.
We developed a new prototype based on just sending reports as e-mail with text descriptions in the body, audio notes and images as attachments. This enabled sending mobile reports to the MAG database with off-the-shelf devices and without loading any kind of software. They could receive the information on the mobile browsers on some of these new mobile devices.
This new approach made the mobile reporting platform more affordable and reliable, however, the interfaces are still very complex. It's hard to record and send audio messages. You can do that on many phones. Most people don't know how to do that, and if they do it takes multiple steps to record your message and send it as a digitised audio file attached to the e-mail.
We created a database that would record the mobile reports with the date, time, location and report description, and we had categories. We described fields for the type of place or service, and the accessibility feature being reported on. The type of place could be a restaurant, bank, shop, church. Service types could be transportation, health care, equipment repair. Accessibility features including things such as curb cuts, entrance, seating, restrooms.
We conducted a demonstration trial at the 2005 RESNA conference last year in Atlanta, which is the Rehabilitation Engineering Society of North America. We had 30 camera phones donated by Cingular to use for free over the week of the conference, and we recruited mobile reporters to document accessibility problems and solutions that they found. We had our own team of mobile reporters out there updating and submitting these to our database.
A web page, accessible through a standard browser, presented these mobile reports, and a mobile PDA browser allowed access to the reports on smaller format mobile Internet devices. We tested an iPaq PDA and an iMate PDA with a camera, a tablet PC and laptop and the Cingular camera phones.
The front page of the RESNA news web site presented stories and updates on the conference and events around Atlanta. For example, the story "RESNA comes to Atlanta" described how to get down to the food court. Mobile reporters posted an alert about picketers blocking the street while protesting labour problems. The web site included a variety of the kinds of information we thought would be helpful to people attending the conference and would demonstrate the value of community of interest sharing information.
What we learned was that it didn't work very well. We couldn't get the community of interest to share the reports quickly. Maybe we could have anticipated that, but the volunteers needed more training. We handed out the phones and showed them very quickly how to use them, but it was complex for them to take the pictures and attach them to the e-mails, and to send voice messages. The text reporting was very difficult. We had very few people submitting text reports.
It was also difficult to motivate the participation because the attendees were very busy with other activities. Until you start building enough information to make it useful, people aren't going to go out of their way to access the website or try to look it up on their mobile browser. You need to have a critical mass of information really. While the RESNA demonstration trial didn't work very well, I think it can still succeed when the cell phone interface makes it easier to submit reportsand when a critical mass of reports attract people who will want to regularly contribute information.
Some of the next steps we are looking at now are new location aware devices. One of the problems with the Cingular phones we used was that they had to enter a venue code for their location. The Cingular phone didn't have GPS, and so it was difficult to get people to submit that information.
We are working now with Research in Motion. They have donated Blackberry devices that have GPS built in, so location is automatically tagged to the mobile report. We are also expanding this to include the mobile Google Maps application, where the reports that have been submitted can be presented on a Google Maps interface. We are looking at extending this to other RERC projects, such as wearable captioning and audio wayfinding.
In our work with Google Maps we have integrated the mobile reports with filters and a map display. When a mobile report is submitted, our customised Google Maps page puts a marker where an accessibility report is available. The users can filter reports based on place or service, accessibility, and later on your personal preferences and interests can filter those reports.
In another Wireless RERC project, Leanne West is leading an effort to provide wireless text captioning on a mobile device, either on the device display or as a floating text on a display that's hooked to eye glasses so you get floating text captioning in the eye glass. The idea is to provide captioning in public venues or movie theatres, but using location aware devices you can add content to that and simplify the interface, so it brings up the text captioning for that venue without having to select the venue or select the text channel. Based on your personal preferences and location, the captioning device could also be used to provide public announcements in either text or other formats. We are looking at expanding this to audio comments as well, so audio description, text and sounds in the environment.
I also would like to mention a project that Dr Bruce Walker is leading. Bruce, a professor in the School of Psychology at Georgia Tech, is conducting research on a system for wearable audio navigation (SWAN) in which 3D spatialised audio beacons provide audio waypoints that you can follow.
The system combines location tracking, a way of presenting this audio information in a non-speech format using 3D sonification in head-centred co-ordinates. Customised audio display elements describe different features along the path. For example, different sounds are used to represent curbs, parks, and street signs. Input and control options for the system include speech, a keyboard, a Twiddler device, keypad or braille. For output options, Dr. Walker is looking at headphones versus bonephones that don't interfere with hearing.
The SWAN concept of wayfinding using an auditory display determines a person's location, figures out what's around them and then represents each object with a unique sound.
The listener learns what a location sounds like, so it's not just the waypoints you hear as you follow the path, but audio information describing if a park is nearby or a fountain, or a cluster of stores just the next block over. How do you represent this complex information in audio format is the area of Bruce's research - cognitive perception of sound - and how do you describe the environment using audio effectively.
In the examples of location based research that I have presented today, there are several common needs that we have identified. We need better ways of describing accessibility information in a standard format, so people can use a common language to describe problems and solutions that they encounter. There are many different location based products and applications that could use this standardized information. There are navigation products such as the Sendero GPS or the Trekker, and there are research groups and governments trying to disseminate information to citizens and visitors.
A common language is needed to describe community features so that people can share information. For example, it would be very helpful to define a standard way of describing the information necessary to get through a complex traffic intersection. These standard descriptions could be easily understood by other consumers who share that need and interest. We also need better filters to distribute relevant accessibility information and intelligent algorithms to simplify the interface.
What are the challenges for location based sharing of accessibility information? Better location tracking accuracy, indoors and out, is needed. Better presentation methods are needed on mobile devices because it remains very difficult to see tiny displays in bright sun or to hear mobile devices outside. Battery life is a continuing problem. What information resources should be included in shared databases, and who monitors the community input? That's a big issue as well as privacy. How do you integrate this with existing services and when do accessibility alerts become a nuisance? Again, it gets back to the issue of how you filter that information, so you provide what's relevant but you don't overwhelm people with information they don't want.
In summary, communities of interest can improve safety and access and the joy of exploration by sharing information. We still must seek better ways to present that information in a meaningful and relevant way to people without distracting them and interrupting them as they go about daily life. Thank you very much.
| Contents page |
Last updated: 14.11.2007 © Copyright reserved
