You are here: Home >> Biometrics >> What the future of the automatic teller machine may look like?

What the future of the automatic teller machine may look like?

Wiesław Bicz

Many of the readers of this article have certainly had a chance to use a terminal with help of which transactions are made in shops and restaurants by those customers using credit cards. I do not of course refer to the equipment of the older type, called "an iron," which transferred with a roller - as a stamp - the convex letters onto a pre-printed form, but its modern successors. These are small computers, which read the information encoded in a magnetic strip or a chip contained in the card, posses a keyboard, display, printer and also - this being quite important - can connect with bank computers to check account standing, or to make money transfer. Those devices are usually connected with electronic cashiers, and from the point of view of the payer their handling is done by means of inserting the card to a slot and entering so called PIN. Today it is assumed that the number should be known only to the card's holder and to constitute something, which univocally identifies and authorises a transaction. This replaces in some way a signature placed previously on the pre-printed forms being handled only with the "iron."

Unfortunately, and this is probably nothing new to the readers of this article, such protections do not constitute a significant barrier to offenders, dishonest card's holders or employees of firms issuing such cards. The number of frauds harming cards' holders and the cards' issuing firms constantly increases and it seems that this tendency can be stopped only by a technology, which will allow a 100% guaranteed recognition of the card's holder. Such technologies have been being developed for some time now by various firms and exist under a name of biometrics. Their creators hope that with an assistance of analysis of voice, appearance, pattern of an iris and, primarily, fingerprint lines, they will be able to provide error-proof identification of people. Much points at that that the most practical technology will be the one, which will enable error-proof identification of fingerprints. It has been known for more than a hundred years now that those patterns with no doubts whatever can identify their owners. Unfortunately construction of such a piece of equipment which will have the facility to perform this with equal accuracy to dactylography specialists is not a simple task.

It so happens that patterns of fingerprint lines possess low contrast and the classic methods of their recognition are too sensitive to dirt, water, grease, which by their nature are present on their surface, and also it is not easy to ascertain that the examined object is really a finger - and more so an alive one. In this article I will not attempt to consider the problems connected with that and to discuss the existing ones or the proposed solutions along with their benefits and drawbacks - this would require a longer work. However I am going to present a solution, which has a chance not only to ensure error-free identification of the card's holder, but also to replace the today's terminal altogether.

Such a solution is being developed in Optel, a firm based in Wrocław, which is based on the idea that ultrasonic waves are the best medium for fingerprints recognition, the method having been proposed in 1986 by the author of this article. Construction of the piece of equipment that was based on this proposal was not easy - it required mastering of many new solutions and many years of work. The fruits of this work are prototypes protected with many patents, and - quite constructionally advanced - a vision on what the future version of the terminal may look like (and at the same time the envisaged production scale version of this device). In my opinion this may be of interest to readers of this article. Also because of the fact that such a piece of equipment can be useful not only there where a transaction is to be performed with a credit card, but also it can replace the classic door locks, passwords for computer access, etc. This does not have to be a very complicated device. Its basis will constitute a glass plate, on which edges there will be placed transmitters and receivers of ultrasonic waves. Under the plate there will be located a liquid crystal screen. Apart from that this will also require employment of electronics, which can be located in only a few chips. The whole thing can occupy e.g. a square of 10 cm side (such dimensions will have the first prototype of the device), although there are also possible devices of larger or smaller dimensions.

The whole device is based on the principle that contact of any thing with the plate's surface will effect a change of way of propagation of sonic waves inside of it, which allows observation of that thing which comes into contact with it. It would be difficult to explain in detail in what way the device "sees" what is happening on its surface in such a short article. To provide only a brief explanation, there have been utilised for this purpose ideas known from holography, although in configuration that has not been seen before, and the whole device is nothing else but a holographic acoustical camera - undoubtedly one its kind.

Images being produced by the camera simply copy those places, which come into contact with the plate. Then the fingerprint lines of the pressed-on finger can be seen, but also of course it can be established where it is located. This allows not only its recognition, but also - somewhat by the way - elimination of a keyboard (since the whole plate is an equivalent of today's touch screens and can constitute a virtual keyboard). Of course a pencil can also get into contact with the screen and it can be observed what position it is in. Undoubtedly the pencil can be also used for writing or drawing, and the device will be capable of following its movement - similarly to today's drawing pads. This function can be also utilised for transfer of a signature. For recognition of the user there is also another thing important: because ultrasonic waves partly penetrate the pressed-on object, it is possible to ascertain if the matter is a real skin or a silicon dummy; it is also easy to verify, if the finger is alive, because of visibility of the blood circulating through it. It is also worthwhile to mention that traces left by other fingers - classic fingerprints - remain entirely invisible. Ultrasonic waves - opposite e.g. to light - simply do not respond to them.

To provide the full picture it can be also said that the whole device does not have to be specially expensive, and because it will not have any mobile parts, it can be expected to be very durable.

I suspect that inquisitive readers have already noticed that the device could be used also for one more purpose: as it can be easily noticed the relief of fingerprint lines is a certain, specifically coded information, given to us by the nature and telling something about human genes. One can easily imagine that any given information can be encoded in this way, similarly to the way stamps are made. If this is done on the surface of a credit card, then an ultrasonic terminal could perform to some extend a similar function to the classic "iron" - it will see the relief on the surface and will read the information encoded in this way.