john Gill technology header image

Future Changes in Information and Communication Technology relating to Visually Impaired People

 

John Gill
RNIB Scientific Research Unit

 

Technological developments have helped visually impaired people, but they have also resulted in extra problems.  For instance the increasing use of self-service machines with visual displays, such as cash dispensers and mobile phones, mean that access to services may be restricted for those who cannot read the visual display.

The general approach has been to encourage designers to incorporate features in the standard product which will help people with disabilities.  If this is insufficient, then to incorporate a standard method of connecting the user's own device which has an appropriate user interface.  But if neither of these approaches provides a satisfactory solution, then use special equipment.

Inclusive design is not just adding an extra feature to a product to meet the perceived needs of a disabled user.  It is a process, like quality, which has to be considered at every stage in developing a new product or service.  This requires companies to promote a culture of inclusion within their organisation.  It also requires detailed technical guidelines on the design features required by the various groups of disabled users.

 

Daily Living

Since over half of visually disabled persons in the UK live alone, they have to cope with cooking, housework, taking medication and personal care.  In addition they may want to maintain their garden and join in social activities such as sports.  A variety of ingenious devices have been developed to assist with specific activities.

Developments for the general public are not always to the advantage of visually disabled persons.  For instance, the standardisation of packaging means that aerosol containers of oven cleaner and hairspray feel the same.  One small step was the introduction of an embossed triangle on packaging of dangerous substances.  A few items are specially designed to be easy to differentiate by touch to help blind persons; for instance the bank notes in the UK are of different sizes depending on denomination.

Labels are often in small print with poor visual contrast.  It is somewhat surprising that the labels on eye drop bottles are often difficult to read despite the probability of the user having impaired vision.

There have been a number of proposals for using the barcode to help visually impaired people sort their groceries at home.  The barcode gives the product number so it would be necessary to have a databank to relate this number to the product name or label information.  However it might be possible to have a barcode reader connected to a database via the telephone or internet, with the output being synthetic speech.  The cost of establishing and running such a service has not made this proposal attractive to potential operators.

In the past the controls on most domestic appliances (eg washing machines, cookers and central heating) could be modified for a blind person by adding embossed markings to the control panel.  However the change from electro-mechanical controls to dynamic visual displays has meant that other solutions must be found.  It has not been economically viable to modify each device individually to give speech output. However developments such as Bluetooth may mean that it will be viable to connect an external speech synthesizer to the appliance.

Bluetooth is a radio system for interconnecting systems such as mobile phones, televisions, and central heating controllers.  It works at about 2.4 GHz and has a range of 10 metres (or 100 metres with additional amplifiers).

Accuracy and speed for reading a visual display will be affected by the contrast on the display, which can be adversely affected by ambient light reflected from the screen.  Unfortunately high contrast displays tend to consume more electrical power and cost more than low contrast displays.  Therefore the designers of many consumer products have opted for low contrast displays despite the problems this will cause for many elderly people.

Popular activities such as gardening can pose problems for blind people.  It is most important to remember precisely where one left a gardening tool; one gadget which can help is a device which emits an audible beep when triggered by a signal such as hand clap.

However it has proven harder for blind people to fully integrate with sighted people in sports activities.  Although there are a number of sports in which blind people can participate, most of them require some changes to the rules or equipment to accommodate the blind person.

The telephone has been a boon for many blind people since they have been able to use it with similar ease to a sighted person.  However many of the new services are difficult or impossible to use since they require the user to read a liquid crystal display.

For most of the population, television is a major source of entertainment.  For someone with impaired vision it can be difficult to follow the plot.  Therefore a few programmes, on digital television, are audio described.

Audio description is where a description of the visual aspects of the scene is inserted in the gaps in the dialogue.

 

Mobility & Orientation

The environment in which we live is becoming increasingly complex.  Even journey across a city by bus requires a range of skills including:

These tasks may seem trivial but for someone with no useful vision they are skills which have to be learnt.  Even for someone with low vision, all these tasks are less easy than for someone with normal sight.

The traditional symbol for a blind person is a white stick.  For mobility purposes, this was developed into a long cane which is swept in front of the blind pedestrian to check for a clear path.  Since it is rigid it can transmit information about the surface texture.  The cane has to be light as well as rigid, so aluminium tubing is often used for construction, but carbon fibre technology has been utilised to reduce weight even further.

A significant disadvantage of the long cane is that it provides no information about obstacles at head height, such as lorry wing mirrors or overhanging holly bushes.  This problem can be overcome by the use of a guide dog.  Guide dogs have a number of disadvantages;  that they cannot be taken everywhere and they require care.  For some people, they are the ideal mobility aid but total demand is estimated to be less than 4000 dogs in the UK.

Over the last thirty years, engineers have devoted considerable resources to developing electronic systems to help a blind person avoid obstacles.  The most common approach has been to use ultrasonics; as with radar, the range is obtained from the length of time it takes for a pulse to be reflected back to the transceiver.  Other systems have used lasers or infra-red.

Many of the devices just provided information about the range of the nearest object; a 'picture' could be built up by moving the sensor from side to side.  Other devices have attempted to give a more complete image of the environment but at the expense of providing an excessive amount of information to the blind user.

The main problems are not in designing the electronic circuitry for a satisfactory electronic mobility aid but in:
1.         Identifying the optimum information needed for independent pedestrian travel.
2.         Displaying this information to the blind persons in a non-visual format (usually auditory or vibratory signals).
3.         Manufacturing the device at a reasonable price.
4.         Training blind persons in the use of the device.

The capacities of the senses of hearing and touch are very small compared to that of the visual channel for a human.  Therefore selecting and processing the information to make best use of the non-visual channels is not a simple task.  The sensors in future devices are likely to involve more than one modality (eg both a video camera and an ultrasonic transceiver) in order to obtain the necessary data which can be processed to produce an accurate image of the immediate environment.  However the research which has been done on the automatic processing of satellite pictures and the research on neural networks offer hope that significant advances could be made in the next few years.

For a blind person, the problem of getting about is not just that of not walking into objects.  One problem is that of knowing the layout of the environment; here, an embossed map can help.  However embossed maps are not easy to produce or interpret since just embossing a sighted map seldom leads to an intelligible embossed map.

The problem of converting a sighted graphical representation to an embossed one can be illustrated by the problem of indicating direction.  Visually it is often shown as an arrow on a line.  An embossed arrow gives a sense of direction at only one point on the line and the symbol is unfamiliar to many blind persons.  However a line sawtooth in cross-section has an indication of direction over the whole length of the line, and it is easy to associate the symbol with the meaning since the line is smooth in one direction and rough in the other.

Computer-aided design systems have been developed to speed up the process of producing embossed maps and diagrams.  However there is still much work to be done on the design of the maps and on methods of tactual reading.

Even with an embossed map and a mobility aid, it is still very easy for a blind person to get lost.  A number of electronic aids have been developed, but few have been widely used because of the cost of modifying the environment.

One type of system uses infra-red transmitters mounted at street corners; the infra-red signal is modulated so that a receiver, held by the blind person, gives out an audible message.  These systems can also be used to indicate the status of traffic lights.  Similar radio-based systems have been used in some countries, but the advent of Bluetooth is likely to dramatically reduce the cost of installing such systems.

A different concept is for the blind person to carry a tag similar to the ones used in shop security systems.  Thus machines can detect the presence of a blind person within a few metres and modify their behaviour (eg give out a speech message).  The tag or smart card can be pre-coded, which could indicate that the person would prefer messages in an alternative language.

Satellite navigation systems, such as the American Global Positioning System (GPS), can be used to determine one's position to a few metres.  However this requires line-of-sight to three or four satellites, which means being outdoors and not close to tall buildings.  This position is just given as latitude and longitude, so it needs to be integrated with a detailed digital map of the area.  Just such a system was successfully developed by the MoBIC project.

An alternative method of determining one's position is possible from mobile telephony by determining the relative signal strengths at different base stations.  With the next generation of mobile systems, this has been further developed so that sighted users can be provided with information related to their locality (eg the location of the nearest cash dispenser or Chinese restaurant).  The advantage of the mobile telephony system is that it does not require line of sight to satellites, but the accuracy may not be as good as GPS.  But, as always, the price charged to blind people for the equipment and using the service will be a significant factor in determining its takeup.

Access to Information

It is a myth that vision is a finite commodity and that using it means that it will fail sooner.  When conventional spectacles give insufficient correction, low vision aids can be used.  These devices include simple hand-held magnifiers, stand-mounted magnifiers, and spectacle-mounted and hand-held telescopes.  In general the higher the power of magnification the smaller the field of view and the shorter the working distance (ie the space between the aid and the material to be viewed).

Closed-circuit television systems offer the possibility for the individual to select the appropriate magnification combined with image enhancement (ie improved contrast) and image reversal (eg white print on a black background).

Lighting is probably the single most important factor in facilitating reading by older people.  In general older people benefit from high levels of illumination, but the problem is to obtain these levels without the user suffering from glare.  The illumination on a surface is inversely proportional to the square of the distance from the illuminaire.  Therefore bringing a reading lamp to half the distance will give four times as much light on the reading material.  Many people benefit from simple solutions such as replacing a conventional bulb in a reading lamp with a spotlight bulb of the same wattage; this results in seven times as much illumination on the reading material.

The choice of typeface can significantly affect the legibility of a display. For applications such as interactive television which may be used for sending emails or reading a password for home banking, it is essential that there is clear differentiation between I, l and 1.

One of the most useful aids for a blind person is a tape recorder since it requires few special skills to record or to listen to the material.  However there are a number of problems such as speed of playback and indexing.  If a tape recording is speeded up, the pitch goes up making it difficult to understand (it sounds like Donald Duck).  However research done many years ago by the Royal Navy on communicating with deep sea divers breathing helium led to electronic circuitry to compensate for this frequency shift; the helium gives the divers a very high pitched voice.  There is still no satisfactory solution to indexing on an ordinary compact cassette.

However digital technology offers the possibility of combining text and audio, and incorporating sophisticated searching and indexing facilities.  For reading an audio novel, the main advantages of digital technology would be improved audio quality.  For a cookery book, the new technology could offer facilities such as direct links to nutrition information.  For an academic reference book the possibilities become even greater as long as the publisher had the resources to incorporate the extra facilities.

The future of electronic books for the general population is surrounded by considerable hype.  However American legislation has forced the main developers to consider the needs of visually impaired persons.  So it is hoped that these developments will lead to an increase in accessible literature for blind people.

An important aspect of this digital technology is the ability to transfer data over high speed telecommunications links.  For instance ADSL (asynchronous digital subscriber line) works over ordinary telephone lines but could transfer a typical audio novel in under 10 minutes using MP3 compression techniques; this might be economically attractive if the telecommunication network operators were to offer inexpensive call charges during the night for visually impaired persons.  The precedent for such charging could be the free postage for material for blind people offered by the Post Office.  Third generation mobile telecommunication systems offer speeds up to 2 Mbps when in the vicinity of a transmitter, so this could be another method for delivering talking books to blind individuals.

The best known communication system for blind people is braille which was developed by Louis Braille.  A combination of six dots can only have 64 different configurations.  This gives a problem in that many more than 64 different characters are used in modern print texts; so, at times, more than one braille character is needed to represent one print character.  A braille book is typically 20 times as bulky as the print edition.  Therefore a form of shorthand is employed which uses 189 abbreviations and contractions giving a space saving of about 25%.

For computer use it is becoming common to use an eight dot braille system which can represent 256 characters using only a single cell.

Since there is a shortage of skilled transcribers, computer systems are often used to translate text to contracted braille which is then output on a special embosser.  The algorithms for this translation are not simple since the rules governing the use of contractions depend on pronunciation  and meaning.  For example, there is a contraction for 'mother' which can be used as part of a longer word as long as it does not bridge a syllable boundary as in 'chemotherapy'.

Since one of the greatest deprivations caused by blindness is lack of privacy, the provision of bank statements in braille has been a very popular service for over 25 years.  This system can be totally automated since bank statements are in a fixed format.

Layout is more problematic for mathematics and music.  Since braille mathematics is written on one line, the conversion of the layout on the printed page to a meaningful form in braille is far from trivial.  Braille music is also significantly different in layout from sighted music notation since the braille reader has to read linearly.

There are about 13,000 people in the UK who regularly read braille (NB this should be compared to the one million people whose vision is such that they could be registered as blind or partially sighted).  For these people it is a very useful communication medium since it can be written as well as read by a blind person.  This level of readership, which is typical of developed countries, is partly attributable to the difficulty in learning a new method of communication by people who lose their vision later in life.  Another factor is that diabetic retinopathy is a significant cause of visual disability among those of working age, and diabetes usually adversely affects the sense of touch.

The moon code was developed by Dr William Moon of Brighton in 1847.  Since it has similarities to ordinary print characters, it is easier to learn by blind persons who previously read visually.  However it has the disadvantage that it takes about 80 times the volume of the print version.  Also, the high cost of production has meant that very few books are printed in this medium.  The number of moon readers has dwindled to about 400, most of whom are in the UK.

Up to recently moon was produced boustrophedon (which literally means turning as oxen in plowing) which meant that the user did not have the problem of backtracking from the end of one line to the beginning of the next.  This type of presentation is also called (serpentine) or meander; it was used by the Greeks in ordinary writing at about 0 AD.  The disadvantage was the perceptual problem of reading alternate lines in reverse.

The increasing use of graphics in printed books, particularly school text books, gives problems.  Although many diagrams can be converted to an embossed form, the process of reading by touch means that a diagram has to be tactually scanned and a mental image built of the whole diagram; this is the opposite process to visual reading where one looks at the overall picture and then reads the detail.

Some years ago NASA had a problem with communicating with astronauts during lift-off.  The problem was of information overload using visual and auditory communication.  Therefore they investigated the use of tactual communication; the project failed, but the research formed the basis of a reading aid for blind persons.  The Optacon used a hand-held camera connected to 144 piezo-electric elements which gave a vibratory image of the print character; the task of recognising the character was left to the human.  Its use has been limited by the low reading speed (typically 40 words per minute after extensive training).

Systems to recognise printed characters have been developed for inputting text to computers.  Such systems have immediate application for visually disabled persons since the information can be output in synthetic speech, on a braille display or in large characters on a screen.  None of these systems can read hand-writing satisfactorily.

Recent developments in speech technology have led to significant improvements in quality and accuracy, and a reduction in prices.  This is important for people with disabilities since speech technology can be used for interpersonal communication, access to information and control of the environment.

Although the intelligibility of the available speech synthesis systems is quite high for a number of languages, there is evidence that significantly better quality should be achievable.  Research is focused on naturalness, prosody and flexibility (eg different voices).  Other important research areas are modelling of emotions and speech synthesis from the meaning of the text.

Recent developments in speech recognition have been in two areas - the academic where the focus is on improved techniques such as stochastic modelling, search and neural networks, and the pragmatic where the focus is on accuracy, robustness and computational efficiency permitting real-time performance with affordable hardware.
Speech recognition is difficult because:

Speech dialogue systems, in which speech synthesis and recognition are used in a man-machine dialogue, have been demonstrated for applications such as bank cash dispensers and hands-free voice diallers on public telephones.  Current systems are limited to vocabularies of a few hundred words, but laboratory prototypes exist with vocabularies of a few thousand words.

Speech synthesis is a stable technology, but speech recognition is still substantially inadequate when compared with human capabilities particularly in noisy environments.  Research on speech processing has had a significant impact on the design of cochlea implants as well as digital hearing aids.  However there is a need for better understanding of user needs and matching these needs to appropriate research and development.  There is also a need to use current knowledge in practical situations to evaluate potential benefits.

A fax machine can be used to transmit hand-written text to a central office where a human reads it back over the telephone.  With the decreasing cost of fax machines, this technique is looking increasingly attractive as a means for providing a remote reading service.

In the last twenty years, the most important change for many blind people has been the advent of the personal computer.  With text-based operating systems, such as DOS, a blind person could access information with similar ease to a sighted person.  The output from the computer could be in synthetic speech, a transitory braille display or large characters on the monitor. 

However the introduction of the graphical user interface, of which Windows is the best known example, brought a range of new problems.  Early versions of Windows were partly inaccessible and required the blind person to have an understanding of the structure of the operating system.  More recent versions of Windows have been more accessible because Microsoft has built in more accessibility features, but even so it is still not as easy as DOS for a blind person to use.  Some blind people only use DOS, but most new software is not available in a DOS version.

Email, since it is text-based, has been very useful for many blind people since it is relatively easy to learn and to use.  The world wide web offers exciting possibilities for accessing large quantities of information but there are problems.  Firstly the blind user needs a suitable browser.  The main problem is that many web sites use graphics such that they are not meaningful when accessed by a text-based browser (eg with speech or braille output).  Guidelines have been produced for how to design accessible websites, but these guidelines are widely ignored by commercial organisations.  So some web sites are accessible, but these tend to be ones belonging to government departments.  The popular websites, eg for home shopping and home banking, are still largely inaccessible.

 

Deafblindness

The term deafblind is used here to refer to a person having combined loss of hearing and sight to such degree that he or she cannot make immediate use of facilities for those with impaired hearing or sight alone.  This definition therefore includes people who have a combination of a severe hearing loss and low vision.

The estimates for the numbers of deafblind people varies considerably because very different definitions are used in different countries.  However using the definition above in developed countries, the numbers will be of the order of 200 per million of the population.

For a device to be useful to a deafblind person, it does not necessarily have had to be designed specifically for the deafblind.  However more and more devices for blind people employ audio output, such as synthetic speech, so there are fewer inexpensive devices with tactual output.  For instance the number of electronic calculators with braille output is falling since the cost of synthetic speech output has dramatically reduced in the recent years.  These trends have been to the advantage of blind people but have significantly reduced the choice available to deafblind persons.

Since the number of deafblind persons in a country is numerically small and they are not a homogeneous population, the cost of developing and marketing devices to meet their needs is high.  In the past the non-profit organisations often subsidised both development and manufacture, but these organisations are now under pressure to minimise their financially loss making activities.

imageHowever although there has been a decline in the range of lower price and often low technology devices, the advent of computer systems and the internet has opened up new possibilities for users who can cope with the technology and who can find someone to pay for the purchase and running costs of the equipment.  These developments have also required organisations serving deafblind people to develop training and support skills for these high technology systems.

Communication Devices
The simplest form of communication device is a magnetic board with raised metal letters (as used on notice boards); this has the advantage of simplicity but it is very slow. A faster device is one where a keyboard is connected to a single-cell braille display; the disadvantages are that the deafblind person must be able to read braille, and the cost of the device. Single character displays are difficult to read, so it would be better to display a line of characters but this increases the cost.  There are a number of more sophisticated electronic devices which can be used over the telephone via a modem (a device which converts the digital signals into a form suitable for transmission along a telephone line).

Clocks
These are usually alarm clocks with raised markings on the dial and an electrically activated vibrator for the alarm. For safety reasons, it is preferable to run the vibrator from a low voltage source (such as rechargeable batteries). The vibrators are typically small electric motors with eccentric weights. Clockwork vibrators are not used in any of the standard products; the reason for this is not clear. The vibrator does not need to be directly connected by a wire to the clock - it can be triggered by a low power radio signal. This means that the vibrator can be worn on the body (eg like a wristwatch).

Door Signals
The simplest of these doorbell signallers involve a push-button being connected to an electrical device that then transmits a signal to activate a vibrator worn by the deafblind person. Sometimes the transmission is done via a closed loop aerial; this has a number of disadvantages including the high cost of installing the aerial and that the device only operates in the immediate vicinity of the loop. These systems are being superceded by radio devices. The input to the system is not necessarily a push-button; it is possible to use an infra-red detector, but a simpler method is a pressure pad under the carpet (as used as a component in some burglar alarm systems).

Sound Indicators
A sound indicator is an electronic device that gives out a vibratory signal when it picks up an audio signal (eg a telephone bell) above a pre-set level. Usually the vibratory signal lasts for a fixed minimum time, and the amplitude of the vibration is usually constant (ie independent of the amplitude of the input signal). With some devices it is possible to tune the device to only pick up audio signals at or about a fixed frequency; this is useful in minimising the number of times the device is activated by picking up the wrong audio signal.

Radio Paging and Alarm Systems
The decreasing cost and size of radio paging systems has encouraged the development of modifications to assist deafblind persons. Another significant factor has been the liberalisation of the laws, in many countries, governing the use of very low powered radio transmitters.

Other Low Technology Devices
A small number of devices have been modified to give vibratory output. For example a liquid level indicator, light probe and typing aids. These are numerically very few because the development and manufacturing costs are very high for a small national market.

Tactual Finger Speller
Over the years a number of devices have been developed to emulate finger spelling since many deafblind people do not read braille.  Although these devices work well in a laboratory, there have been problems in making them generally available at affordable prices.

Electronic Mobility Aids
In the past 30 years considerable effort has been devoted to developing electronic devices to assist blind people to avoid obstacles.  These devices use ultrasonic, infra-red or laser range finders, and give auditory or tactual output.  However they have not proved popular among blind users, but some deafblind people have found the devices with tactual output to be useful.

Access to Computers
It was the advent of the personal computer with braille or magnified visual output that opened up opportunities for a significant increase in access to information for deafblind people.  Software for producing large characters on the monitor is relatively inexpensive, but braille displays have remained expensive.

The DOS text-based operating system is easier for many deafblind people than ones, such as Windows, which use a graphical user interface.  However keeping to DOS restricts the choice of software in that most new software is written for the Windows environment.

Fortunately email systems are predominantly text-based and therefore relatively easy to use with a braille display.  The world wide web is more problematic in that many sites employ graphical representations without an adequate text alternative.  The sites which are fully accessible tend to be ones belonging to government departments, and the ones which are largely inaccessible are the popular home shopping sites.  Even with these restrictions the internet has the potential for significantly increasing access to information by deafblind people.  What is needed is a range of affordable user friendly terminals which provide access for deafblind people who need non-visual output but who do not read braille.

Telecommunications
The basic mobile telephone has been of limited use to deafblind people.  However the introduction of Universal Mobile Telecommunications System (UMTS) offers exciting possibilities if the services are affordable.  For instance the ability to transmit pictures of where you are to a service centre and receive textual replies in braille, could greatly assist a deafblind pedestrian in an unfamiliar environment.

Related to developments in telecommunications will be radio-based systems for short-range interconnection of both domestic equipment and public terminals.  Systems such as Bluetooth have the potential to facilitate the connection of assistive devices to a whole range of equipment.  If this technology lives up to the publicity, then deafblind people can anticipate significant improvements for those who wish to live independently.

Broadcasting
However not all technological developments are to the advantage of deafblind people.  With analogue television it is possible to obtain braille output of teletext which gives basic access to the news.  However digital teletext is graphically based so obtaining braille output is much more difficult.

 

Although technology could greatly increase access to information and the ability of deafblind people to participate in society, this is unlikely to happen unless there is a significant and ongoing investment in developing new devices and systems and making them available at affordable prices.

 

Inclusive Design

Increasingly blind people need to be able to use equipment designed for the general public; this includes ticket selling machines at unmanned railway stations, cash dispensers, and public telephones.  In the forseeable future, inability to use such systems is likely to increase the divide between the blind and sighted population; this could include next generation mobile phones, interactive television and electronic purses.

Therefore it is essential that equipment for use by the general public is designed to be accessible by as many people as is reasonably possible.  With the increasing ageing population, this must include people with presbyopia as well as people with a combination of different impairments.

Public Terminals
The solutions to some of the problems of people with disabilities may appear trivial to a non-disabled person, but they can nevertheless have a major effect on the usability of a piece of equipment or access to a service.   For instance many people would like a notch in the facia of the terminal so that they can lean their walking stick against the terminal without the stick falling over.  Other problems require more complex modifications, but often solutions are available but not implemented. 

For many disabled and elderly users, the most important aspect is consistency in the user interface of public terminals; this is particularly important for visually, intellectually and cognitively impaired users.  A prime example of this is the lack of a single standard relating to the layout of numeric keypads.  With public terminals, the user may only use it occasionally and has probably been provided with minimal training in the use of the terminal.  What is “logical” to the average user may be different from what is “logical” to the designer, so it is essential to test any new user interface with a cross-section of potential users (including disabled and elderly people).

To select a preferred interface such as audio instructions or large characters on the screen, the user could simply press a button or otherwise select from a menu on the screen; this is likely to increase the time taken to undertake a transaction if there are more than a few options.  Another possibility is to store the user’s preferences on a central computer and implement them as soon as the PIN (personal identification number) has been entered.

For card-operated terminals, it is possible to store the information on the user’s card (the coding of user requirements is specified in the European standard EN1332-4), and this is in many ways more desirable than storing private information about a user on a central database.  With a magnetic stripe card there is very limited spare capacity for storing this information (but this method has been used for storing the user’s preference for displayed language), but a smart card (containing an electronic chip) has fewer restrictions on storage capacity so appears to be ideal for this purpose.

Many disabled users would like to be able to select and store their preferred interface whenever they use their card at a public access terminal.  It is essential that information is stored on a card only with the consent of the user.

For a blind person, it can be difficult to find the terminal if they are not very familiar with the environment.  One possibility is to use a contactless smart card, carried by the blind person, to trigger an audible signal from the terminal at a distance of a few metres.

Public access terminals can incorporate audio prompts in the form of ‘beeps’, to indicate an action. It is recommended that new equipment should provide a more sophisticated solution of using audio leadthrough in the form of a verbal set of instructions.  Audio leadthrough can assist people with visual or cognitive impairments (and first time users). Message content should be chosen very carefully since a message that might be acceptable to the users for the first few times they hear it may become unacceptable when they hear it for the hundredth time. Many users with impaired hearing, who can only hear lower frequencies, can more easily hear a male voice than a female one.

If audio output is used to provide private information to the user, then it should be through a telephone handset located at the terminal or through a headset connected through a standard mini jack to the terminal; however, it is essential that the position of the jack socket is standardised.  If a handset is provided, inductive coupling and amplification should also be incorporated.

Braille instructions on outdoor terminals have limited value in cold weather since tactual sensitivity is dramatically reduced with decreasing temperature.  The estimated number of braille readers in Europe is less than 200,000 so although useful for some blind users, braille is not a total solution for visually impaired users.

Blind persons, and many elderly persons, have problems in inserting the card in the correct orientation; this is a particular problem on cards which are not embossed.  However there is a European standard for an orientation notch (EN1332-2) in the card:

People who wear bifocals find it difficult to read the screen of most public access terminals, since neither lens gives a focussed image for the distance between their eyes and the screen.  In addition many people leave their spectacles in the car or do not wear them in public.  So the number of people who have problems in reading the screen is much more than those considered “blind” or “low vision”, who constitute about 1.5% of the population.

People with low vision should not be prevented from getting their faces close to the screen.  However it is possible to increase the size of the characters on the screen for individual customers who require this facility.  This can be done by selecting this option from a menu or preferably by storing this information on the customer’s card.  With touchscreen systems, it could be arranged that holding one’s finger in the top left corner for at least two seconds indicates that one would like double size characters on the screen.

Ambient light, such as from an illuminated sign above the terminal, can cause problems if it results in glare or reflections from the screen.

Moving text on a screen can be very difficult to read for someone with even a mild sight impairment, so it should be avoided whenever possible.

Digitally stored speech can give very good audio quality, but it is effectively limited to pre-stored messages.  Full vocabulary synthetic speech is often difficult to understand for the naïve user, particularly if they have a hearing impairment.  Non-confidential information can be output on a loudspeaker, but the volume should be a function of the current ambient noise level; this is less of a problem with handsets or headphones. 

One technological possibility would be for a disabled user to have a hand control unit with an infra-red or radio link to the terminal.  It may be that Bluetooth becomes the standard interface, but still there is a problem in persuading service providers to fit it to all terminals (which would include retrofitting to existing terminals).

Displayed text should use simple, large, bold fonts in upper and lower case characters. Displayed messages should be simple in sentence structure, use natural language, and any graphical symbols (such as icons) should be accompanied by text.

Information, which is sensitive and private to the cardholder, shouldnot be visible to any other persons; screen filters, which act like a slatted blind and restrict the user to be directly in front of the screen, improve privacy but often at the expense of visual quality.  However the user may wish to display information with large character size, but they should be made aware of the privacy problem.

Standard layout of keypads is essential for visually disabled people and highly desirable for other users.  To help blind persons, there should be a single raised dot on the number five key.  However this does not solve the problem of there being two common layouts for the numeric keys (ie the telephone and the calculator layouts); it is recommended that the telephone layout is used exclusively on public access terminals.

Ideally keys should be internally illuminated when the terminal is waiting for input from that keypad.  There should be some form of feedback on key input (eg a beep and/or tactual indication).  Tactile feedback can be provided by a gradual increase in the force, followed by a sharp decrease in the force required to actuate the key, and a subsequent increase in force beyond this point for cushioning.

Many elderly people and those with a cognitive impairment do not like to be rushed or to think that they are likely to be “timed out” by the machine, so it is necessary to allow for such people to use the terminal at their own pace; this requirement could be stored on the user’s card.

To help elderly people and those with hand tremors, key fields should be as large as possible and separated by a ‘dead area’.  There should be high contrast between touch areas, text and background colour.  Avoid using a pretty picture as background - it is a menace to anyone with poor vision or someone reading the screen under difficult conditions (eg in bright sunlight).

For blind users, one possibility is to arrange that holding one’s finger in a specified corner of the screen for at least two seconds initiates speech output (NB this must be a different corner than the one used to request large characters on the display), or tapping twice in the corner.  Another method would be to store this requirement on the user’s card.

Touch screens can either be triggered by insertion or withdrawal of the fingertip.  With the latter system, it is technically possible for the user to pass their fingertip over the screen and get speech output describing the active area they are touching at the time.  Then the system is only triggered by withdrawing the fingertip from over an active area.

To aid visually impaired users, receipts should have a minimum font size of 12 point with a sans serif typeface with upper and lower case text, but 16 point would be preferable if space permits.  It is important that the print has good contrast on opaque paper with a minimum of background pattern.  A common complaint is poor print quality on receipts which can be a result of the printer ribbon not being replaced regularly.

Electronic purses
Over the years there have been many proposals for systems to reduce consumer reliance on cash and cheques especially for low value purchases.  The advent of smart card technology has offered a secure mechanism at an affordable price.  One of the benefits of smart cards is the creation of new value-added services like electronic purses which store pre-paid monetary value directly on the card.

The UK Banking Code defines an electronic purse as "Any card or function of a card which contains real value in the form of electronic money which someone has paid for in advance, and which can be reloaded with further funds and which can be used for a range of purposes".

For some time pre-payment cards have been in general use for applications such as public telephones.  These disposable cards are loaded with a fixed amount of value, which was then decremented during use.  The next step was to make the cards reloadable which can be done at specialised terminals or automated teller machines (ATMs).  Then it is only a small step for the cards to be used for more than a single application.

However these systems vary significantly depending on the type of organisation operating the service.  Historically the banking organisations have placed great importance on security, whereas public transport operators have been more concerned with the time taken to complete the transaction.

In public transport systems there is often a requirement to store centrally a complete record of the transactions.  In part this is motivated by the requirement to accurately allocate subsidies among various operators for carrying passengers who do not pay full fares (eg in some areas disabled passengers are not charged for using local public transport).
Such a central record means that if a card is reported as lost, the card can be electronically cancelled the next time someone attempts to use it, and the rightful owner can be reimbursed with the value on the card at the time it was stolen.  In practice, the central system is often offline and it is updated overnight; this means that there can be up to 24 hours during which the stolen card can still be used.  In other types of system a record of recent transactions is stored on the card and the terminal receiving payment may record the serial number of the card as well as the amount of the transaction. 

The development of electronic purses has been driven by commercial and technological organisations rather than by demand from consumers.

As smart cards contain increasingly powerful microprocessors and large memories, a future development will be to have multi-functional cards with an operating system and a number of independent application modules.  Therefore there is the possibility of a user having a card onto which applications are loaded by individual service providers (eg at an ATM).  However despite the technological viability of such multi-application cards, their introduction is likely to be held up by a myriad of problems such as branding and retailer acceptance.

To use the electronic purse, the user hands the card to the shop assistant who inserts the card in a terminal and keys in the amount of the transaction which is displayed visually to the customer.  The customer confirms that the amount is correct, and the money is transferred from the card to the terminal.  In some systems, the customer needs to key in their PIN (personal identification number) before the transaction can be completed.

In the future, PIN systems may be superceded by biometric methods such as iris patterns, fingerprints or facial recognition.  Ideally users should be able to choose to use a PIN instead of a biometric method if they have problems with the biometric system (eg fingerprint recognition requires the user to have fingers).
Many electronic purse systems provide users with a balance reader the size of a chocolate bourbon biscuit.  These readers tend to have low contrast visual displays which are very difficult to read by people with impaired vision.
Systems, such as Mondex, offer customers the possibility of using an electronic 'wallet' to verify balances and to transfer money from one card to another.  For instance a taxi driver might have an electronic wallet so that he or she can accept electronic payments.  Unfortunately some wallets have a numeric keypad laid out in the telephone format and some in the calculator format.  These wallets are not easy to use by the uninitiated and pose particular difficulties for those who have difficulty in reading the liquid crystal display.  The wallet offers the possibility of 'locking' the card using a four digit PIN.
Some electronic purse systems offer the facility to transfer funds between the card and the customer's bank account using a screen phone.  Screen phones can be modified to be accessible by blind people, so this method of loading a card could be the preferred mode for many blind and partially sighted users.

For visually impaired persons, the main problems with electronic purse systems relate to the user interface.  In particular many of the devices (eg balance readers, wallets) have poor contrast visual displays which also pose problems for many elderly persons.  One possibility would be to have special versions of the balance readers and wallets for those who cannot read the standard version.

Such a special wallet / balance reader might include:
(a)        High contrast display with larger characters (or speech output).
(b)        Larger buttons which have clear visual markings and tactual feedback.
(c)        A funnel opening to help guide the card into the reader.
(d)       Instructions in large print.

One disadvantage of this approach is that too many of the general public might want the special version.
At the retailer's premises it is important that the visual display is positioned such that the customer can easily read it before agreeing to the completion of the transaction.  For customers with low vision, this requires that they can get close to the high-contrast visual display.  If possible the display should be white characters on a black background; red displays are often used for technical reasons, but some people with low vision have problems with this end of the spectrum.  Another possibility would be for there to be speech output of the transaction amount; if this was not wanted for all customers, this requirement could be coded on the user's card.

Mobile Phones
Mobile telephones increasingly require the user to read a small liquid crystal screen to operate many of the functions in the phone.  Although the phones incorporate increasingly powerful microprocessors, manufacturers have not seen a commercial opportunity in providing models which incorporate speech output of the messages normally displayed on the screen.  However there are some indications that this may change because car drivers are seen as a significant market segment.

The SIM (subscriber identification module) card in a GSM (global system for mobile communications - the digital mobile system) phone can be used for providing interaction with internet and other services.  This will not involve true browsing, but will include 'push services' such as checking train times, booking theatre seats and restaurant tables, traffic news services and purchasing.

Considerably more functionality will be available with WAP (wireless application protocol) which will facilitate financial transactions such as reloading an electronic purse.  Digital signatures and encryption can reside on the phone itself, in the SIM chip, so there is no need for an additional smart card for authentication of transactions.  However problems in agreeing a common standard for the smart cards may mean that the phone will need two card readers.

Extra functionality to suit visually impaired users could be built into the terminals, but this in itself is unlikely to provide full access to services.  Therefore it will be necessary to modify the server or proxy server.  The WAP User Agent Profile Specification covers aspects of the technical interface (eg the technical capabilities of the terminal) and the User Preference Profile concerns content selection (eg the user is interested in receiving sports scores).  Neither of these profiles covers the needs of people with disabilities.

This could be done in the form of a user profile which is stored on the smart card in the user's phone.  There is already a European standard, but this will need to be extended to allow for the facilities needed by mobile phone users who have disabilities.  This standard already incorporates facilities for specifying preferred text size, screen colour, colour avoidance, voice output, and interface complexity.

GPRS (general packet radio service) is a high speed packet data technology which will permit data transmission speeds of up to 100 kbps over the GSM network.  This is well suited for frequent transmission of small amounts of data.  However it could be overtaken by UMTS.

UMTS (universal mobile telecommunications system) is the next generation, after GSM, of mobile telecommunications system which will have seamless operation between terrestial and satellite links.  It will provide high speed access to the internet with data rates of up to 2 Mbps for a stationary terminal, less when on the move.  It will include packet data transmission with the potential to adjust bandwidth on demand for asymmetric traffic.  In the UK the network licenses have been sold by the government for vast sums (the first five licenses sold for over £22.5 billion) which will have to recouped from the users.

A major problem is the desire of the regulatory bodies to keep the network completely separate from the terminals.  For many disabled users, what is important is the end-to-end connection and whether it meets their needs.  In many cases, it is going to require a combination of facilities in the network and on the terminal, and there will be difficulties in getting universal service for people with disabilities while there is the regulatory separation of network and terminals.

Digital televison
Interactive television is attracting considerable investment as it is seen as a significant step in selling new services to customers who may not be computer users.  The UK government envisages that it will be a major method of interaction between the public and the government within five years.

Traditional television can be characterized as one-to-many whereas the internet would be characterized as many-to-one.  As yet interactive television is frequently just an enhanced one-to-many system with uncertainty as to how to satisfactorily also be a many-to-one system.

Various companies are already offering interactive television services in the UK, and many more are planning to launch services in the next year.  The services vary considerably depending on the main service being delivered:

 

Smart housing
The interconnection of devices in the home has been held back by the lack of a consensus on appropriate standards, but this may be partly resolved in the foreseeable future by such systems as Bluetooth..  However the television set may become the central display for such a system, and therefore it is essential that allowance is made by the designers for people who have difficulty in reading the screen or understanding the process.

A simple smart house might give an audible warning of any windows left open when someone locks the front door from the outside.  Also it could provide warnings about a cooker left on.  In addition it is possible to connect the internal system to a telecommunications link to provide remote assistance or telemedicine.

The reason for the great interest in this area is the realisation that with an increasing elderly population and, at the same time, a decrease in the number of people working and paying taxes, it will be essential for as many people as possible to live independently for as long as possible.  The cost of the technology involved in smart housing is modest when compared with the cost of residential care.

 

Future Research

Over the next ten years, there are likely to be improvements in the medical prevention and treatment of eye conditions, but the increase in the ageing population will probably mean that the net effect is that there will be a larger number of people with visual impairments.

Many older visually impaired persons have seen little benefit from the advances in the area of assistive technology.  If they use assistive technology, beyond conventional low vision aids, it will usually be the low-technology low-priced devices to assist in daily living.

An area of research which appears promising is vision enhancing systems in which the user is presented with a display with enhanced contrast of the objects in the near field.  Such a system would use a body mounted camera and a head-up display.  Also such a system could be used to partly compensate for loss of colour discrimination.

Research is being done on connecting a video camera to the human brain.  One approach was to implant a number of electrodes in the brain; this proved problematic, so more recent research has concentrated on methods of stimulating the optic nerve.  This research is important but it is unlikely to produce systems which are widely available within the next ten years.

Vision substitution involves converting a video signal to a non-visual form which can be auditory, tactual or some combination of the two.  Since both the auditory and tactual channels have far lower information capacity than the visual channel, it is necessary to process the signal from the video camera before it is sent to the non-visual display.  Ideally this processing should include edge detection and object recognition.  With increasingly powerful wearable computers, vision substitution systems look very promising if the appropriate algorithms can be developed.

Universities discourage research on 'simple' devices because there appears to be little academic content.  Industry often does not perceive there to be a significant commercial market.  This means that this area has been left largely to the non-profit organisations most of whom have concentrated on marketing their existing products.

What is needed are studies of the problems faced by visually impaired people, and the development of novel inexpensive devices to meet these needs.  This will require imagination as well as a knowledge of appropriate production processes.  Last, but not least, all this needs to be linked to a marketing strategy which reaches all the potential consumers.  For too long, visually impaired people have had to make do with products which are poorly designed.

Another aspect is the desirability of developing products which are attractive to non-disabled users.  It is worth noting that the fountain pen, typewriter and long-playing record were originally developed for use by blind people.

There is a particular need for devices for people with more than one impairment.  The problem is that this is a far from a homogeneous group, so it is likely to involve developing devices which are adaptable.

Many research workers find it easier to study the needs of people who are totally blind since it is difficult to do an accurate comprehensive functional assessment of someone with low vision.  This has been particularly noticeable in the area of research on electronic mobility devices.  There has also been a tendency to look for purely technological solutions to human problems.

It is in the area of orientation systems that exciting but practical developments can be expected in the next few years.  However it will be important that service delivery, including training, is developed at the same time.

A major research area relates to graphical representations.  The trend for systems for sighted people is for the increasing use of graphics, but research on non-visual presentations is lagging far behind.  For people with low vision, more research is needed on how to optimally present visual graphics; this may involve transposing colours for people who are colour blind.

The world wide web is partly accessible for users who have to rely on non-visual output.  However new modes of presentation and the rate of development could mean that access could get more difficult.  This is a serious problem since the web has the potential for significantly narrowing the gap between blind and sighted people regarding access to information.

There is a shortage of research workers experienced in working with blind and visually impaired people.  What is needed is a number of research teams with a long term commitment to this area.  A typical team would consist of three post-doctoral researchers, two or three research students, and technical and secretarial support staff.  To establish such teams would require a change in strategy for funding research in this area.

Virtual Reality
In the longer term, technological developments in other fields could have benefit for those working with visually disabled persons.  For instance, virtual reality technology could provide realistic simulations of visual defects including fields, acuity, colour discrimination and effects of illumination;  in addition, multiple handicaps could be simulated.  Such a system might be useful for evaluating proposed public buildings.

Virtual reality technology is of interest since it allows people with sensory disabilities to perceive what they might not otherwise be able to since it can gather information in a sensory modality in which they are impaired and deliver it to one where they are not.  It also has the potential to be used as a training aid for skills such as spatial co-ordination and orientation.

As intriguing as virtual reality is, the enabling technology is still crude.  Major technological hurdles exist in the area of tracking a person’s motion and position in a non-intrusive way, in displaying high definition stereo colour images of the scene covering the user’s peripheral vision and, in the area of image generation, speed for a smooth and realistic animation of the scene.  Tactile output and the construction of physical images, to support the visual images in virtual environments, require further development to produce realistic sensations.

Inclusive Design
The "inclusive design" message has had limited practical impact upon the area of information and communication systems and services.  This is despite considerable effort being expended by various groups around Europe. 

In the case of cash dispensers, the companies manufacturing the equipment see their customers as the banks purchasing their equipment.  Even though they may have incorporated inclusive design features in their range of terminals, it is to no avail if the bank is not interested in offering it to their customers.  Within the bank it may be a technical department which is responsible for selecting equipment for the bank, but it will be the local branches who have direct contact with disabled customers and who may provide a modicum of training in the use of the cash dispenser.  Unfortunately local branch staff are unlikely to be aware of the technological possibilities for improving the accessibility of the equipment.

At the policy level it may be sufficient to specify that the equipment and services must be accessible to as many people as is reasonably possible.  However this leaves open many questions including what does "accessible" mean?  Also what is "reasonable"?  Also it does not cover the often crucial question as to who pays for any additional costs such as training.

The development of guidelines for inclusive design of systems and services in the area of information and communication technologies is seriously hampered by the sparcity of sound scientific data about the needs of people with disabilities.  What data exists is all too often based on inadequate sample sizes or inappropriate methodology.  This is another area which is perceived to be low on academic content.

Industry wants guidelines to be pan-disability, but this will require greater collaboration between all the relevant organisations representing the different disability groups.

However with new equipment and services which are only in the early stages of specification, such as UMTS, it is difficult to be more precise.  However if the influencing is left to the stage when it is clear what features will be incorporated, it is often too late to get anything significant changed.

Information for product designers may be detailed design guidelines (eg the maximum height and angle of a display so that it can be read by a wheelchair user).  However this approach is only possible for established technology for which detailed design guidelines exist.  In other cases it will be necessary to provide generic guidelines backed up by recommendations on how to test prototypes with a cross-section of potential users.  For telecommunication designers the problems are shortage of time and lack of an established system for evaluating with disabled users.  This is an area where user organisations could take a more active role in providing speedy evaluation of prototype systems and services.

The development time in telecommunications has been decreasing which means that the time between a project being proposed and the specification finalised is short.  Also secrecy is considered essential by many commercial organisations.  Therefore the possibility of having direct contact with the product specifier at the right moment is remote.  So, in practice, it is essential to provide the information, or a signpost to it (eg a web address), in advance and hope that the recipient remembers it at the relevant moment.  This can be assisted by the company having a design checklist which includes questions on the accessibility of the product by disabled users.

Policy documents from the European Commission are written in a special language which is difficult to understand by the uninitiated; only recently have some of the organisations representing disabled people taken on staff with the skills to interpret these documents.  However these organisations frequently do not have the technical expertise to understand some of the implications.  Therefore there needs to be some form of collaboration between those who understand the language and the regulatory issues, those who have a good grasp of the technology, and those with lobbying skills.

Standardisation
Standards are crucial in the telecommunications industry where there is a rigorous, but if sometimes slow, process for developing standards.  In the television industry, the process is somewhat different in that the technical standards are frequently determined by bodies made up of only industry representatives and there is no policy for involving consumers.  The situation is different again in computer software where the commercially dominant players set the de facto standards with apparently no consultative process.  This means that convergence is going to involve a clash of cultures as well as the more obvious problems of integrating three different groups of technology.

Standardisation is slow, time consuming, lacks academic content and sometimes has limited direct commercial benefit.  Academics shun the area since it does not produce research publications.  Industry is hesitant about inclusive design standards work since they can see no short-term commercial benefit.

Standardisation has four phases:

Industry and many government departments feel it is the role of the organisations of disabled people to assist with the implementation and awareness phases.  However the disability organisations have often taken the view that they should be paid to do this work.

Since governments now see the primary purpose of standards as facilitating trade, the role of the consumer has become somewhat uncomfortable.  Therefore legislation or mandatory regulation would appear to be the only practical way of requiring commercial organisations to adopt inclusive design.  However the current trend is towards minimising regulatory control, and European legislation does not appear likely in the near future.

Over the next ten years there are exciting possibilities for improving the quality of life for people with a visual impairment, but there are hurdles to be overcome to ensure that the appropriate research is undertaken and the results of this research reach all the potential consumers.

 



John Gill Technology Limited Footer
John Gill Technology Limited Footer