Section 10.3. Biometrics and Public Technology: The ATM Example

10.3. Biometrics and Public Technology: The ATM Example

As different application areas place different requirements on biometrics technology, we must consider the appropriateness of different technologies for a specific application. This section explores the applicability of the different biometrics technologies for use within an ATM or other self-service environment.

For a number of reasons (to be enumerated in the following sections), I assume here that verification rather than identification will be utilized at an ATM.

10.3.1. ATM Fingerprint Verification

Integration of fingerprint sensors into an ATM for verification is feasible, especially if contact silicon (capacitance or thermal swipe) devices are used (see Figure 10-1). Optical sensors and static capacitance systems are not viable for ATMs because of dirt and latent prints that accumulate with use. Swipe devices, on the other hand, at least have the potential for self-cleaning with use. Unfortunately, the physical positioning of the sensor may be difficult for all users (left-handed/right-handed aside) to get their fingers flat on the reader. Smart cards equipped with fingerprint sensorswhere the fingerprint is used to allow use of a private keycould remove many of the technical problems of deploying biometrics in the self-service environment.

Figure 10-1. Fingerprint reader

10.3.2. ATM Face Verification

Another possibility for ATMs is the integration of facial biometrics systems (see Figure 10-2). In many cases, a security camera is already present; this camera could be used for both video recording and biometrics verification.

Because face verification could be a universal biometricseverybody has a facemost research on this technology to date has focused on the potential negatives. Several complicating factors have been identified:

• The system's field of view must cover a wide range of heights, from the tallest standing user to a user in a wheelchair.

• Today's technology requires that users find and look at the camera at the right time, facing straight on and with a neutral expression.

• Lighting must be uniform and consistentwith good front lighting and little back lightingposing a significant problem in many locations.

• Some systems can be tricked into accepting photographs or even drawings of faces.

• It is possible that people could attempt to defraud the system through decapitation of a legitimate user. The use of masks is another avenue for potential fraud.

• Likewise, an attacker could coerce a legitimate user to look at a camera to complete an authentication under duress.

• There are unknown accuracy rates, with a high chance of false negatives.

On the positive side, face recognition presents no hygiene-related health concerns and no current association with criminology. This biometrics technique can also be used without the user's cooperation or even his knowledge, as a backup or addition to any other biometrics deployed, potentially raising privacy concerns.

Figure 10-2. Face recognition

10.3.3. ATM Iris Verification

Iris verification is the most reliable biometrics, but currently it is still very expensive for systems that do not require the user to put her eye in close proximity to the camera (see Figure 10-3). (Such positioning could cause problems in an ATM environment for the same reason that it is problematic for face recognition.) Iris verification has already been piloted in ATMs.

On the positive side:

• The method is noncontact and noninvasive.

• It can be located anywhere.

• The physical attribute it uses is relatively stable and it is a proven system in the ATM field.

On the negative side:

• The risk of fraud by the use of force is possible.

• Criminals who are not familiar with how the technology works could attempt fraud through mutilation of the user. Although this would not work, as a responsive eye is required, the mere thought of such a crime is exceedingly distasteful and could impact adoption or use.

• The technology requires more user cooperation than face recognition would.

If performance accuracy were the only criterion for adoption, this system would be more popular. However, the difficulty of positioning the eye remains a problem. Advances in both algorithms and robotics to direct the camera to the eye instead of requiring the user to position the camera or eye manually will surely make this technology more successful.

Figure 10-3. Iris scanner

10.3.4. ATM Retina Verification

Many reasons exist for not using retina verification and existing retina systems in ATMs. These include:

• The technique is invasive.

• The technique requires training and patience to use.

• Although the system requires no physical contact, users must still place their eyes very close to it, making sensor location difficult to accommodate all users.

• Users must look into the output from a laser.

• Questions exist as to the stability of the measured feature (body temperature and fatigue, for example, have been quoted as possible causes of variance within the pattern).

Retina verification systems may evolve, and theoretically their performance is attractive, but at present they have too many negative aspects to deploy in any capacity.

10.3.5. ATM Hand Verification

The finger geometry system has some potential, but at present the integration of a hand sensor into an ATM would be difficult for a number of reasons:

• The size of the larger systems can be prohibitive (9 cm (W) x 15 cm (H) x 14 cm (D)).

• As with all optical devices, dirt can pose a problem in self-service environments.

• Positioning of the sensor may be difficult for all users (left-handed/right-handed aside) to get their hands flat on the plate or their fingers in place.

• Fraud through the use of severed hands may or may not work. Forcing someone to present his or her hand is also a possibility.

• Vandalism or accidental damage of the units could be a problem.

10.3.6. ATM Speaker Verification

Speaker verification uses the user's voice as an authenticator. There are many advantages to this technique at an ATM:

• It involves simple integration of the necessary voice-recognition sensors.

• The sensor need not be at the ATM at all. For example, a mobile phone could be used.

• The system is more resistant against coercion, because it is more difficult to force someone to speak than to physically move that person to the location of the sensor.

• In the event of coercion, the stress of forced speech could alter the voice sufficiently that the user would no longer authenticate.

However, there are some negatives:

• Background noisefor example, aircraft flying overhead or the noise of rainstormscan cause serious difficulties. (Handsets would ease this problem but would create others.)

• Enrollment is time consuming (unless mobile phones are part of the system).

• Template size may create problems not associated with other biometrics techniques.

• Recording of the human voice may create a simple fraud potential, although some companies claim recorder detection.

10.3.7. ATM Signature Verification

With a signature verification system, the computer verifies a physical signature that the user writes with either pen-and-ink or a special pen on an electronic pad (see Figure 10-4). This system is very attractive for use with financial transactions because of the long historical and present-day association between signature and financial authorization. What's more, an image of the signature can be preserved for later human scrutiny.

Signature verification systems can be integrated directly into an ATM, although the location of small tablets and pens can pose a usability problem for some users.

Although this technology is intriguing, it has not yet proven to be reliable.

10.3.8. ATM Typing Verification

It is unlikely that everyone would have sufficient typing skills or practice to be comfortable with the typing verification technique, even if a standard alphanumeric keyboard were available on all ATMs. Such a keyboard would require space and careful placement so that its use could mimic that of a desktop-based keyboard. If full keyboards were to become the norm on many self-service devices, one could see the attraction of this relatively unknown biometrics technique, but otherwise, the technique is infeasible.

Figure 10-4. Signature verification