2004 Conference Proceedings

Go to previous article 
Go to next article 
Return to 2004 Table of Contents 


AFB TECH PRODUCT EVALUATIONS: INSULIN PUMPS, CELL PHONES, AND VOTING MACHINES

Presenters
Mark Uslan, Managing Director
American Foundation for the Blind
Technology and Employment Center at Huntington (AFB TECH)
949 3rd Ave., Ste 200
Huntington, WV 25701
Phone: (304) 523-8651
Email: muslan@afb.net

Darren Burton, National Program Associate in Technology
American Foundation for the Blind
Technology and Employment Center at Huntington (AFB TECH)
949 3rd Ave., Ste 200
Huntington, WV 25701
Phone: (304) 523-8651
Email: dburton@afb.net

Craig Swisher, Intern
American Foundation for the Blind
Technology and Employment Center at Huntington (AFB TECH)
949 3rd Ave., Ste 200
Huntington, WV 25701
Phone: (304) 523-8651
Email: afblabhun@afb.net

The American Foundation for the Blind Technology and Employment Center at Huntington (AFB TECH) evaluates the accessibility of products for use by blind and visually impaired persons. This presentation will report on findings to date on three product evaluation projects: Insulin pumps, cell phones, and voting machines.

Insulin Pumps

Of the 17 million people with diabetes in the U.S., approximately 5.3 million have diabetic retinopathy, the major cause of blindness among people with diabetes. The use of insulin pumps (IPs) has revolutionized diabetes care because it more closely mimics the normal pancreas than multiple insulin injections. Use of an insulin pump can lead to a freer lifestyle, improved flexibility in meals, the ability to exercise or travel without risk of losing blood sugar control, and fewer diabetes-related health problems. From 1990-2001, the usage of insulin pumps increased exponentially from almost 9,000 to over 160,000.

To test the feasibility of and barriers to their use, 9 IPs were evaluated for accessible use by people who are blind or visually impaired. The pumps evaluated were from Medtronic Minimed, Sooil, Animas, Disetronic, Deltec, and Nipro. Features and functions such as: operating procedures, user interface design, menu structure and menu navigation, and computer interface capabilities were tabulated and analyzed. Particular attention was paid to the accessibility of safety features and the control of the delivery of insulin. In order to identify potential failures and set design priorities, a Failure Modes and Effects Analysis (FMEA) of insulin pumps was conducted.

Results indicated that none of the 9 or so commercially available IPs were designed to be fully accessible to blind and visually impaired persons. It was determined that the following readily achievable improvements can be made to the basic user interface of insulin pumps to make them significantly more accessible to people who are blind or visually impaired:

  1. Developing an audio output for all IP menus.
  2. Improving tactile differentiation of IP interface buttons
  3. Improving infusion set design so that it is easily connected and disconnected by blind and visually impaired persons.
  4. From the FMEA analysis, focus accessibility design improvements on pump priming, insulin bolus delivery, and insulin basal rate delivery.

Although designing insulin pumps to be fully accessible to people who are blind or visually impaired should be the ultimate goal for IP manufacturers, the technical hurdles to achieve this goal are formidable. Ultimately full accessibility will likely await the development of a fully integrated, closed loop insulin pump/blood glucose monitoring system that requires minimal user input.

Cell Phones

As cell phones have become an increasingly conspicuous part of everyday life, they have also become more and more powerful, equipped with computer operating systems and more features being added with every new model. Web browsers, e-mail, multimedia messaging, voice dialing, phone books, word processors, multi-color displays and cameras are some of the features available on today's cell phones, but there are concerns among the blind and visually impaired community as to whether or not these and other cell phone features will be accessible. Section 255 of the Communications Act, as amended by the Telecommunications Act of 1996, requires that cell phone manufacturers and service providers do all that is "readily achievable" to make each product or service accessible.

To determine the accessibility of today's cell phones for people who are blind or visually impaired, AFB TECH evaluated 6 of the top cell phones on the market in 2003. The phones evaluated were the Audiovox 9500, the Motorola T720, the Sanyo 5300, the Sony-Ericsson T68i, the Nokia 3650, and the Nokia 9290 Communicator, which is actually a phone/personal digital assistant. The 2 Nokia phones were evaluated equipped with 2 European speech output software products: Mobile Accessibility developed by Code Factory, and TALKS from Brand & Grober.

To focus our evaluation work, we first created a questionnaire to survey people who are blind or visually impaired to determine which features they would most like to be made accessible. This resulted in the "Sweet 16," a list of 16 features that we then evaluated to determine whether or not a blind or visually impaired person could use them and noted the barriers to using them. Evaluation methods included measuring the ability to tactually identify keys, determining the ability to navigate menus, assessing the readability of the displays, and noting auditory and vibratory feedback.

Results indicated that the 4 phones without speech output software had very few accessible features, mainly due to the lack of access to screen information. Furthermore, 3 of the phones had such tactually unidentifiable keys that it was extremely difficult to do even the basic tasks of answering and placing phone calls. Although the phones with the speech software loaded on them certainly give blind and visually impaired users access to many more features than the other phones, they still do not provide access to all the features that sighted users have. Additionally, the cost of these phone/software combinations ranges from $500 to nearly $1000, putting them out of the price range of most people. However, they do show that it is possible to provide accessibility via speech software, providing hope for the future of cell phone accessibility.

Voting Machines

The chaos surrounding the 2000 presidential election clearly illustrated the need to modernize our nation's election systems, and several manufacturers have developed electronic voting machines to eliminate the inaccuracies involved with paper and chad-style ballots. Since it is estimated that 500,000 of these machines, costing between $2,000 and $5,000, would need to be purchased across the United States, it could be a multi-billion-dollar process to bring our nation up to date. The Help America Vote Act (HAVA), passed by Congress and signed into law in 2003 provides federal money to states to purchase at least one accessible electronic voting machine in every polling place by a January 2006 deadline. Having accessible voting machines would give blind and visually impaired voters the ability to cast a private ballot independently, without having to rely on pollworkers or friends or family members to cast ballots for them.

To help ensure that people who are blind or visually impaired are not forgotten as this modernization occurs, we gathered five machines that are on the market and evaluated their usability and accessibility. The machines were the iVotronic from Election Systems & Software, the AVC Edge from Sequoia Voting Systems, the eSlate from Hart InterCivic, the AccuVote TS from Global Diebold, and the Vote-Trakker from Avante International Technology.

Fifteen people, ranging in age from 18 to 64, volunteered to act as testers to evaluate the voting machines. They were observed and filmed while using the machines and then interviewed afterwards. These tests provided data to evaluate the following accessibility features:

  1. Speech quality and whether the speech is produced via human voice recordings or synthetic speech.
  2. Clarity of both printed and spoken instructions.
  3. Controls that are identifiable tactilely and/or have proper braille labels.
  4. A means of avoiding undervoting or overvoting.
  5. User control of font size and screen contrast.
  6. The ability to use visual and audio voting simultaneously.
  7. Overall ease of use of the system.

The evaluation results showed that although there is certainly room for improvement, all of the machines were in fact accessible, providing speech output and keyboard input to complete the voting process. However, ease of use was determined to be the most important feature, because these machines must be usable by all voters, even those whose experience with technology goes no further than the telephone. Also, there is no opportunity to practice with these machines, so they must be easy to use without much training or instructions. We like the eSlate the best because of its overall ease of use owing to its linear ballot style. The runner-up is the iVotronic, edging out the AVC Edge because of clearer speech output and a lesser tendency to cause confusion. Had the AVC Edge's help system not led to the confusion mentioned previously, it would have edged out the iVotronic. Trailing the pack is the Vote-Trakker. Although we found it to be an accessible and usable machine, it loses favor because of difficulties that less technology-savvy voters may encounter with its synthetic speech and computer keyboard interface.


Go to previous article 
Go to next article 
Return to 2004 Table of Contents 
Return to Table of Proceedings


Reprinted with author(s) permission. Author(s) retain copyright.