john Gill technology header image

Selecting Cards by Touch

Dr John Gill
Hannah Devine-Wright
March 1999

ISBN 1 86048 019 5
© Copyright reserved, 1999


For a blind person, or someone suffering from impaired vision, tactile markings are not just useful, they are essential to enable that person to identify items by touch. On behalf of all the blind and visually impaired people of Europe, I welcome this report.

It calls on card issuers to incorporate embossed symbols on all cards to help people differentiate one card from another. This is absolutely necessary when we have to deal with so many cards. It explains appropriate ISO and CEN standards and proposes a new CEN standard. This will meet the needs of those having to operate self-service terminals, giving them the independence and privacy which they need and deserve.

I commend the report and congratulate its authors on a job well done.

John Wall
European Blind Union


1 Introduction

2 Investigation

2.1 Stage One
2.2 Stage Two
2.3 Stage Three

3 Discussion

4 Recommendations

5 References



A - Moon Code 15

B - Symbols Used in the First Phase of Testing 16

C - Visual test of Acuity Used in the First and Second Phase of Testing 17

D - Symbols Used in the Second Phase of Testing 18

E - Proposal for Dimensions and Location of Embossed Symbols for ID-1 Cards 19

1 Introduction

As electronic services become more common place, the likelihood increases of needing to use a plastic card to access these services. In the foreseeable future it is anticipated that there will an increasing number of services which require the user to insert a card in a self-service terminal, and an increasing proportion of these cards are likely to be smart cards with no embossing. People who are elderly or disabled, particularly those people with a visual disability, can find it difficult to select the correct plastic card (eg bank card, telephone card, utility pre-payment card) from their wallet or handbag for use in self-service terminals, particularly where there is a low level of illumination. This is especially problematic when there is no embossing on the card.

The degree to which plastic cards can be visually or tactually differentiated depends to a large extent upon the surface characteristics of the card, notably the size and legibility of characters. The characters on the surface of the card impart information to both the card users and the card reading equipment. However, the effectiveness of printed information on plastic cards is reduced for some users such as people who are elderly, disabled, blind or visually impaired, who find it difficult to distinguish between cards with different functions such as a bank, telephone, pre-payment or social security card.

It is estimated that, in developed countries, about 1 per thousand people have to rely on non-visual acquisition of information and a further 13 per thousand people have low vision. Most visual disability is acquired later in life; about two thirds of people with a visual disability are over 75 years of age. Macular degeneration accounts for about half the visual disability in developed countries; typically this will result in a loss of central vision. For those under the age of 60, the most common form of visual disability is diabetic retinopathy that is often accompanied by a poor sense of touch.

It is not practical to suggest that Braille letters should be used to differentiate plastic cards because Braille readership is low, estimated at 0.02% of the population. Furthermore, Braille is more likely to be learnt by people whose onset of visual disability occurs early in their lives so people whose vision deteriorates as they get older are less likely to be familiar with Braille. Alternative tactile alphabets include ‘Moon code’ that consists of simplified and enlarged versions of the Latin alphabet.

Given the increasing use of non-embossed smart cards and the problems experienced by people with a visual disability in differentiating between plastic cards, it has been recommended that a standard be developed for differentiating plastic cards by touch (Gill, 1997).

2 Investigation

This research sought to develop design solutions to the problem of card differentiation. This was achieved through an investigation of existing data pertaining to the physical design of plastic cards, the canvassing of the opinions and views of suppliers and manufacturers of plastic cards and the testing of possible design solutions with a sample of blind, partially sighted and sighted persons. The outcome of the testing phases informed our recommendations for the design of embossing on plastic cards.

2.1 Stage One

According to EN ISO/IEC 7810:1996 the nominal dimensions of ID-1 cards are as follows:

Width: 85.60 mm (3.370 inches)

Height: 53.98 mm (2.125 inches)

Thickness: 0.76 mm (0.030 inches)

According to BS EN ISO/IEC 7811-1:1996 the maximum height of embossed characters is 4.32 mm (0.170 in). Characters should be spaced 3.63 mm ± 0.15 mm (0.143 in ± 0.006 in). The height of visually and machine readable characters on embossed cards are 0.48 mm (0.018 in).

According to BS EN ISO/IEC 7811-2:1996 the magnetic stripe area occupies a position along the length of the card and is situated close to (max. 5.54 mm/ 0.218 in from) a long edge. The position of a magnetic stripe was not relevant when considering the location of embossed symbols on contact or contact-less ID-1 cards that did not incorporate a magnetic stripe.

According to BS EN ISO/IEC 7811-3:1996 embossing is located within two areas. The first contains visual and machine readable information and the second area is intended to contain information about the card holder eg name, address and is intended to be read visually. The parameters of the ‘name and address area’ are as follows: maximum height is 14.53 mm (0.572 in) with a bottom margin of minimum 2.41 mm (0.095 in) and a maximum of 3.30 mm (0.130 in). However, if a magnetic stripe is present the minimum margin is increased to 2.54 mm (0.100 in). The characters should be situated between 7.65 ± 0.25 mm (0.301 ± 0.010 in) and 66.04 ± 0.76 mm (2.600 ± 0.030 in) of the left edge.

The area available for inclusion of embossed symbols for differentiation of plastic cards will depend upon the amount of information provided on the particular application.


2.2 Stage Two

Institutional and technical limitations were addressed through the distribution of 65 questionnaires to four groups of respondents. Firstly, card manufacturers; secondly, card suppliers, ie for telecommunication, transport or identification, (eg the Post Office); thirdly, interest groups including blind organisations, local authorities and fourthly, others such as designers, electrical retailers or personal commentators. Fifty-two questionnaires were returned with detailed comments. The responses of members of the four groups are summarised in Table 1.

Table 1: Type, number and summary of responses by group membership

Type of respondents

No. of respondents

Summary of responses

Card manufacturers


Against embossing (cost and technology); notches okay; prefer smart cards or personalisation by user



Mixed reaction to embossing: problematic (cost) but refer to existing design guidelines

Interest groups


Emphasis upon generic identification and orientation (embossed, cut or symbol) for users



Consult retailers; generic identification preferred




Respondents raised nine issues:

1) Card orientation may be assisted by increasing tactile discriminability,

2) Visual impairment will not be a disadvantage with the advent of smart cards that do not have visual discriminators on them,

3) Who is responsible for ensuring cards are discriminable: the manufacturers, suppliers or users?

4) Is personalisation viable, eg could individual stickers be applied in a designated area to assist differentiation or could use be made of card wallets?

5) Differentiation should be generic, eg function rather than specific, eg brand,

6) Any solution must accord with existing standards,

7) Colour (high contrast) and sound (at terminal) should be considered as additional discriminating factors,

8) Avoid Braille due to its limited use,

9) Consider international generalisability.

With regard to technical issues, it was noted that various existing standards were referred to eg ISO 7816, ISO 7811. The response to questions about the position of symbols was various. Of the six respondents who expressed a preference, five preferred the top (right or left). Concern was expressed about not interfering with existing information on the cards including avoiding logos, magnetic strips, electronic chips, etc. It was suggested that the height of embossing should be as recommended by Gill (1997) with colours being bright and of a high contrast.

A further questionnaire developed and distributed to 70 suppliers of plastic cards, through the Journal of Healthcare, resulted in a very low response rate. Since, the issues that were raised were similar to those attested to by card suppliers; the few responses were incorporated into the discussion of the views of manufacturers. A central issue to emerge was concern about who should be responsible for the embossing of plastic cards? Manufacturers, who focussed upon smart card technology, regarded embossing on smart cards as unnecessary and costly. It was suggested that current suppliers and users could amend their designs to incorporate embossed, cut or symbolic solutions. However, the position of company logos and electronic chips would limit the position of any proposed embossing.


2.3 Stage Three

Having established the physical and technical limitations of embossing plastic cards it was necessary to investigate possible design solutions. The research for this project was carried out in two phases: each phase utilised twelve different symbols, six of which were common to both phases. Although the position of the symbols in each phase remained constant, their size was varied with the second phase requiring the tactile identification of smaller symbols.


2.3.1 Phase I

Choice of Symbols

The choice and size of symbols was based upon a review of existing literature concerned with tactile discriminability of symbols. Discriminability is increased if symbols are easily recognised and respondents can name the symbols, eg letters are better identified than random, novel shapes (James and Gill, 1974; Lambert and Lederman, 1989). Furthermore, the greater the degree of association between a chosen symbol and the features it is supposed to represent, the easier it is to identify and remember (Bentzen, 1996). Of course, at this stage in the research the emphasis was upon differentiation rather than association. Association was addressed when the final decision was made about the choice of symbols and their allocation to particular card types.

The majority of the symbols used in this research were modified version of Moon code (Appendix A), that was established in 1845. Moon attributes a raised symbol to each of the 26 letters of the alphabet as well as a form for frequently used expressions such as ‘and’ and ‘the’. Despite the longevity of Moon, the discriminability of the symbols has not been formally tested. Nevertheless, the size and shape of Moon code symbols informed the present research. The maximum height of a Moon symbol is 7 mm whilst the width is 7 mm. The depth to which Moon is embossed depends upon the way in which the symbols are manufactured and is not prescribed.

In the present study, the materials consisted of 12 cards each embossed with a single figure (Appendix B) measuring 10 mm by 10 mm. A further 18 cards were embossed with two of these figures and 4 cards were embossed with 3 figures.


Degree of visual acuity was assessed prior to the presentation of the test materials by asking the respondent; "What is your sight like?". In the event that the respondent had some sight, a simple test of visual acuity was administered that consisted of the participant reporting which (if any) of five words or short phrases of different sizes in bold print they could read. Each of the words had been culled from the headlines of a national newspaper (Appendix C).

Each of the single symbol cards was presented to the participant who was asked to judge how easy each figure was to identify tactually using a three-point scale from not at all easy (1), to somewhat (2) or very easy (3) to identify. In all cases, the two and three figure cards were presented after the single figure cards. The order of presentation for each card type was randomised for each participant.

Participants in phase I

A total of ten people were presented with the plastic cards: five women and five men. Of these, four people were sighted (although they did not use their eyes to study the cards), one person was partially sighted and five were blind. One blind participant was familiar with Moon code and one with Braille. The sighted participants constituted a convenience sample. Otherwise, participants were drawn from local blind and partially sighted societies. The age of participants ranged from 27 to 83 years with an average age of 55 years.

Results of phase I

An index of tactile discriminability was calculated by adding together individual responses (minimum 1 to maximum 3) for each of the 12 symbols to give a total score. The minimum score was 10 (ie all respondents rated the symbol as not at all easy to identify) and the maximum score was 30 (ie all 10 respondents rated the symbol as very easy to identify).

Irrespective of degree of visual acuity, the participants found the majority (ie 8 of the 12 symbols) very easy to distinguish (index score was more than 27). However, three of the symbols were difficult to identify tactually, ie the vertical and horizontal ‘M’ and the modified ‘P’ (index score less than 22). Frequently, the modified ‘P’ was felt to incorporate a curved rather than a straight vertical line especially amongst the blind subjects. The ‘B’ shape was considered to be moderately identifiable (index score equalled 24).

Although rated as easy to identify by most people, the ‘X’ and ‘+’ symbols were reported to be somewhat interchangeable by two respondents. Overall, tactile discriminability was improved by the symbol having a clearly defined central rather than peripheral shape. For example, the squared ‘U’, the ‘V’ in any orientation and the ‘O’ shape were easy to identify.

In order to investigate the role of age and degree of visual acuity upon responses, a total score was calculated based upon the cumulative rating of all symbols for each respondent. As there were 12 symbols, each respondent could score a maximum of 36 (ie rate each symbol as very easy to identify). It was noted that the lowest scores were associated with the two (older) sighted respondents (who did not use their eyes but only scored 27 and 29 out of a possible 36 respectively). Meanwhile, the highest rate was afforded to a similarly aged person (59 years as opposed to 54 and 53 years for the sighted respondents) who was totally blind and reported some loss in tactile sensitivity due to employment in heavy industry. It was not possible to identify any simple relationship between age and visual acuity across respondents and knowledge of either Braille or Moon (limited to two people) did not significantly improve performance in this phase of the testing.

The results of the multi-symbol testing replicated the single symbol presentations. The ease with which a particular symbol could be identified singularly was similar when presented as one of a pair or triumvirate of symbols. For this reason, the results of this testing are not presented in detail. However, the results suggested that degree of tactile sensitivity and size of fingers (especially for men) were more likely than age or sight to influence response rates especially during the discrimination of two or more symbols. In the case of two or more adjacent symbols, a space of at least 2 mm is recommended to assist differentiation.


During this phase, it was noted that blind participants read the symbols with their thumb whilst sighted participants read the symbols with their index finger. According to the blind participants, they would use their thumbs to read plastic cards in situ since it would be more convenient (faster and less obvious to onlookers). It was important for the symbols to be quickly identified using a thumb rather than an index finger and to be read at an angle since this replicated how the cards would be used, eg when removed from a wallet.

Overall, the response to the majority of the symbols was very positive. This was reflected in the request by some blind respondents that the symbols be made available to them in a ‘sticker’ format. In this way, they could be used at home to identify the function of household appliances, eg on the washing machine, the microwave or to differentiate between the sides of audio tapes and the content of CD’s.


2.3.2 Phase II

The test materials consisted of 12 symbols that replicated the maximum size of numerals on existing plastic (bank) cards, ie 4 mm by 5 mm. The selection of 12 symbols included rotated versions of the most easily discriminated symbols used in phase I (Appendix D). Each symbol was positioned as in phase I.


The procedure for phase II was identical to that used in phase I. Degree of visual acuity was assessed using printed words of different sizes as necessary (Appendix C). Each participant rated ease of tactile discriminability of each symbol (random order of presentation) on a three-point scale ranging from not at all easy (1), somewhat (2) through to very easy (3) to discriminate. Responses were recorded along with basic sociodemographic information and any additional comments or observations made in the process of conducting the testing. Following testing, each participant was presented with examples of the materials used in phase I upon which they commented (usually about the comparative ease with which they could tactually identify them although this information was not systematically recorded).

Participants in phase II

Fourteen people were presented with the 12 cards singularly. All the participants were accessed during meetings arranged by local blind and partially sighted societies. Eleven of the respondents were female and one respondent was male. Their ages ranged from 20 to 94 years with an average age of 66 years. However, this average was skewed by the inclusion of one 20 year old. A more representative statistic would be that the modal age was 76 years. Five participants were totally blind and the remaining seven were partially sighted. One person could read Braille and one other participant was familiar with Moon.

Results of phase II

Given that there were 14 participants, the cumulative index for each symbol was between 14 (minimum) and 42 (maximum). The index was considered high if the cumulative score was more than 38, medium between 30-37 and low when 29 or below. Accordingly, five of the symbols were highly discriminable (four of which were smaller versions of the symbols used in phase I). Four symbols scored between 30-37 and the remaining three symbols were considered to be difficult to discriminate. The three most problematic symbols were: the inverted ‘V’, the ‘X’ and the ‘Z’.

There was no appreciable effect of age on performance. However, it was noted that the lowest score of differentiation (16 out of a possible 36) was afforded to the eldest participant (94 years), whilst the youngest participant (20 years) correctly identified all 12 symbols and therefore gained a maximum score of 36. The performance of the younger participants may have been assisted by the increased likelihood that they were familiar with tactile alphabets such as Braille having been introduced to these systems at blind schools. A sex effect was immeasurable given the predominance of female participants. Degree of visual acuity (blind or partially sighted) did not seem to influence performance.

3 Discussion

Given the ease with which respondents identified the ‘V’ in any orientation in phase I, it was surprising that the inverted ‘V’ in phase II was difficult to differentiate. Given the method used to construct the test materials, it was considered likely that in this case, the inverted ‘V’ would have been more easily identified if the reproduction in plastic had been clearer. It should be noted that since the ‘X’ was difficult to identify in both phase I and phase II, this could be problematic for its future usage on social security benefit cards that are being introduced in collaboration with Post Offices.

The most easily distinguished symbols (aside from the ‘O’) incorporated straight lines with angles of more than 45 degrees (preferably 90 degrees). For example, the [ or ] in any orientation or the –, T or  symbol.

4 Recommendations

1. ID-1 cards should be embossed with symbols to aid differentiation.

2. Symbols should be a maximum of 10 mm by 10 mm. In terms of manufacturing it may only be feasible to emboss to a size of 7 mm by 7 mm.

3. The use of embossed symbols will assist generic differentiation by function.

4. It is recommended that embossing be carried out on contact or contact-less cards within the area currently specified for the embossing of account and name of holder details.

5. Emphasis should be placed upon modifying plastic cards at the manufacturing or supply stage.

6. Appendix E should be used as the basis for a CEN standard.

7. An appropriate mechanism for service providers to register their symbols should be established.

8. If and when a standard is finalised, publicity should be planned to make potential users aware of the presence of embossed symbols on plastic cards.

5 References

Bentzen, B. L. (1996) Choosing symbols for tactile maps. Journal of Visual Impairment and Blindness, 90(2), 157-158.

Gill, J. M. (1997) Differentiating plastic cards: is there a need for a standard? London: RNIB Scientific Research Unit.

Gill, J. M. (1998) Access prohibited? information for designers of public access terminals. [accessed 06/09/07].

James, G. A. & Gill, J. M. (1974) Mobility maps for the visually handicapped: study of learning and retention of raised symbols. American Foundation for the Blind Research Bulletin, 27, 87-98.

Lambert, L. M. & Lederman, S. J. (1989) An evaluation of the legibility and meaningfulness of potential map symbols. Journal of Visual Impairment and Blindness, 83, 397-403.

Appendix A

Moon code

Appendix B

Symbols used in the first phase of testing

Appendix C

Visual test of acuity used in the first and second phase of testing

Appendix D

Symbols used in the second phase of testing

Appendix E

Proposal for Dimensions and Location of Embossed Symbols for ID-1 Cards


Increasing use is being made of machine readable plastic cards. However, some potential user groups such as people who are elderly, disabled, blind or visually impaired experience a notable degree of difficulty in using existing card layouts to distinguish between cards with different functions such as a bank, telephone, pre-payment or social security card. This has led to the recommendation of a standard for differentiating plastic cards (Gill, 1997) since the task of differentiating between machine readable cards can be a major obstacle to their use. This problem can be reduced by increasing the visual and tactual discriminability of plastic cards, ie by modifying the surface characteristics of the card, notably the size and legibility of characters. Since the surface characteristics of the card are multi-functional, imparting information to card reading equipment and card users it is important to ensure that any modifications do not impede the functioning of the card. Any recommendations are made in the knowledge that good design for disabled persons should be good design for everyone.

Specifically, this proposal addresses the needs of users who have difficulty reading non-embossed information on plastic cards, including persons with special needs, for example older people, minors, disabled people, visually impaired people, those with learning difficulties, first time users, and those not conversant with the local language. It is based on the premise that embossing ID-1 cards with symbols would aid differentiation between cards. It is proposed that embossed symbols would assist generic differentiation by function. Provision for the incorporation of up to four symbols on each card would permit a degree of specific differentiation, ie by card type in addition to card function.

It is recommended that embossing be carried out on contact or contact-less cards within the area currently specified for the embossing of account and name of holder details. Although this proposal does not prescribe exactly who should be responsible for embossing plastic cards with tactile symbols, it is suggested that modification of ID-1 cards be carried out at the manufacturing or supply stage so that the maximum number of consumers can benefit from the inclusion of tactile signifiers.

The aim of this proposal is to specify the design principles necessary for embossed symbols to be incorporated into machine readable cards. The scope of this standard would be limited to ID-1 cards without magnetic strips, ie contact or contactless smart cards. The location of tactile symbols has been informed by existing guidelines/ standards pertaining to the location of embossing on machine readable cards and the standards that specify the location of electronic chips. The recommendations have been informed by a series of tests for discriminability of the symbols amongst blind, sighted and visually impaired persons. Tactile symbols are only required on cards that can not be identified tactually by other means.


1 Scope

This proposed standard specifies the form, dimensions and location of tactile identifiers for application with ID-1 cards. These symbols would appear at the option of the card issuer.


2 Normative references

This proposed standard incorporates, by dated or undated reference, provisions from other publications. These normative references are cited at the appropriate places in the text and the publications listed hereafter. For dated references, subsequent amendments to or revisions of any of these publications apply to this proposed standard only when incorporated in it by amendment or revision. For undated references the latest edition of the publication referred to applies.

EN 742:1993 - Identification card systems - Intersector ID-1 card location of contacts for cards and devices in Europe

EN ISO/IEC 7810:1996 - Identification cards - Physical characteristics

BS EN ISO/IEC 7811-1:1996 - Identification cards - Recording technique - Part 1: Embossing

BS EN ISO/IEC 7811-3:1996 - Identification cards - Recording technique - Part 3: Location of embossed characters on ID-1 cards

3 Definitions

For the purposes of this proposed standard, the following definitions apply:

3:1 ID-1 (Identification card): Card identifying its holder and the issuer which may carry data required as input for the intended use of the card and for transactions based thereon.

3:2 back: The face of the card that is not embossed.

3:3 front: The face of the card which may carry embossing (see figure in EN 27810). [EN 742:1993]

3:4 machine readable cards: Cards incorporating a technology such as a magnetic stripe or integrated circuit that may be read by a machine.

3:5 embossing: To raise characters in relief from the front surface of a card.


4 Abbreviations

IC - Integrated circuit

ID-1 - Card conforming to EN ISO/IEC 7810 card type ID-1


5 Requirements for standard embossed symbols

The following are the most desirable requirements for the design of embossed symbols that will act as tactile discriminators on ID-1 cards:

  • clearly distinguishable by touch: the tactile symbols should be easy to feel and recognise by users including elderly, blind and visually impaired people;
  • clearly visible: the tactile symbols shall be easy to see and recognise by users who are not blind;
  • consistent position: the tactile symbols shall be positioned such that the card can be visually or tactually read with ease;
  • user testing: the tactile symbols should be selected on the basis of user testing to verify that the design meets the requirements of the intended user groups;
  • compliance with existing standards: a standard for common tactile symbols for ID-1 cards shall comply in all respects with existing standards for such cards;
  • machine reading compatibility: the tactile symbols should not interfere with the functioning of existing card reading technologies, eg embossing, magnetic stripes and integrated circuit cards (Figure 1);
  • mechanical compatibility: the tactile symbols should be designed and positioned so that they will not interfere with the proper working of the mechanical card reading technologies, eg card feeding machines (Figures 1 and 2).


6 Dimensions of visually and tactually readable embossed symbols

These symbols are not intended to be machine readable.

Symbol set: Simplified alphanumeric, capital characters and symbols.

B Banking

E Utilities

X Social security

T Telephone

| Information

+ Credit

- Debit

Ñ Health/medical

< Communication

> Transport

O Personal identification

L Loyalty/membership

] (additional category)

(additional category)


Most of the symbols can be reproduced in various rotations (90, 180 or 270 degrees) in order to maximise the use of characters and minimise the cost of implementation. The most easily distinguished symbols (aside from ‘O’) incorporate straight lines with angles of more than 45 degrees.

Symbol to denote absence: A single dot, ' ', placed in the centre of the area reserved for embossing with symbols shall denote an absence of a specific character.

Symbol spacing: The symbol spacing shall be 5 mm ± 0.1 mm.

Symbol height: Maximum height at the printing surface for the embossed symbols, encompassing centreline skew and character misalignment shall be 7 mm.

Note that the maximum permitted height of embossed letters is currently 4.3 mm.

Symbol width: Maximum width at the printing surface for the embossed symbols shall be 7 mm ± 0.1 mm.

Relief height of embossing: The height of visually and machine readable embossed symbols shall be between 0.45 mm and 0.48 mm.

Coding system: A maximum of four adjacent symbols shall be positioned in the embossing area (Figures 1 and 2).

The symbol positioned on the left-hand side of the front face of the card should describe the generic function of the card: banking, social security, health, personal identification, etc. Each subsequent symbol to the right of the first symbol should indicate the specific function of the card (position A in Figure 2). In the event that a card is multifunctional, emphasis should be placed upon use of the generic identifiers. Future categories of card function are permitted by the use of rotated versions of symbols prescribed in this document. When one or more of the four embossing positions are not required to contain a tactile symbol, a ' ' should be embossed to denote the absence of any other symbol.


7 Testing for conformance with this standard

The provision of tactile symbols does not require specialist measurement procedures to test for card conformance with this standard.


8 Labelling and packaging

No special labelling or packaging is required by this standard.

Figure 1: Area for embossing
Measurements in mm

Figure2: Details of embossed area
Measurements in mm


John Gill Technology Limited Footer
John Gill Technology Limited Footer