Automatic identification systems are a broad class of devices that are used primarily in commercial settings for security authorization and inventory/logistics control. Although different devices may employ radically different technologies, they are united in the common purpose of collecting and tracking data about people or objects. Familiar examples of automatic identification systems are bar codes, magnetic stripes, and shoplifting deterrent tags. The companies and trade groups responsible for developing and implementing such devices are often known collectively as the automatic identification and data capture (AIDC) industry.
Automatic identification systems afford many advantages to businesses. They offer a much faster alternative to manual data entry and verification, and a well-designed system is much more accurate as well. Perhaps the simplest example is at the supermarket checkout counter. If product manufacturers and supermarkets didn't use bar codes, cashiers would have to manually read a code from each item or a list, and key the code into the register. If the cashier misread or mistyped, he would have to try again. This is, of course, what was done before bar code systems were implemented, and it is easy to recognize the efficiencies gained by automating that process.
Beyond efficient data collection, automatic identification systems perform tasks that would be much more difficult to accomplish manually. Magnetic ink character recognition (MICR), for instance, allows merchants to detect whether a customer is attempting to pass a counterfeit check. In addition, many applications of automatic identification are more sophisticated and create greater efficiencies than simple bar coding on consumer products. These include quality control in manufacturing, freight tracking, and biometric and voice recognition for security purposes.
Bar coding is the most widely used automatic identification technology. Bar codes began to enjoy popularity in the early 1970s with the advent of the Universal Product Code (UPC) for grocery systems. A bar code is a combination of printed bars and spaces representing letters or numbers. According to the Automatic Identification Manufacturers (AIM), an industry trade group, more than 250 different bar code structures exist, but only about a dozen are in widespread use.
A traditional linear bar code such as the UPC includes a start and stop character and a parity or check character. These characters enable the bar code to become a self-contained identification label. While bar codes essentially signal presence or absence of print, they are more properly considered a type of font. A font characterizes or categorizes printed alphabetical or numerical characters based on size and style. Because bar codes contain lines and spaces of varying widths, they resemble font specifications. The smallest width of any bar code is called the X dimension or, alternatively, the module width. Bar codes are also frequently multidirectional: they can be read in either direction, top to bottom or bottom to top, left to right or right to left.
Two-dimensional (2-D) bar codes are a subject of increasing interest to manufacturers and other businesses that require more sophisticated data storage and retrieval than a conventional linear bar code provides. While the newer 2-D codes are commonly labeled as bar codes, they, in fact, don't necessarily use bars at all. Instead, they may employ a matrix of shapes and spaces resembling a pixilated, or highly magnified, black-and-white computer image with a number of geometric shapes scattered throughout. In the technical jargon, these representations are termed "symbologies," and if they appear random to the human eye, they are very meaningful when scanned into an appropriate automatic identification system. Other 2-D codes simply stack multiple linear bar codes on top of one another. The main advantage of 2-D systems is in the large volume of information they can encode in a small space: a single code can store as many as 4,200 alpha-numeric characters, the equivalent of 700 words.
Magnetic stripes on credit cards, driver's licenses, mass-transit tickets, and numerous other objects function by storing digitally information about the card—and sometimes its user—for retrieval by an electronic reader. The AIM reported in 1998 that more than 20 billion such cards are used each year. Magnetic stripes can be used to reduce data entry, track information, and establish the authenticity of the card in question.
Dating back to the 1960s, magnetic-stripe technology usually involves placing relatively weak magnetic codes on one or more tracks along the magnetic stripe. Although many stripes serve as read-only data sources, generally the stripes can be recoded at any time given the proper equipment. The weakness of the magnetic charge on many consumer cards makes them vulnerable to accidental damage from exposure to magnetic fields. More durable magnetic encoding is available, but because it is more expensive it is usually reserved for more critical applications—those in which the costs of having a bad stripe outweigh the higher material and encoding equipment costs.
So-called smart cards may be used in ways similar to magnetic stripe cards, storing personal identification and financial account information, but the underlying technology is quite different and the functions are more diverse. Widely carried in Europe, smart cards contain an embedded microchip with memory for storage, and more advanced ones also have on board a microprocessor that allows the card to make decisions and encrypt data in different ways. This storage and processing power makes smart cards more secure and more versatile than magnetic stripe technology. Among the data that can be stored are images of an individual's face or fingerprint for identification purposes.
Frequently, three zones of information contain the smart card database: secret zone, a private zone, and a public zone. Public zones track transactional information. Private zones contain user identified passwords. Secret zones contain information from the card issuer that can provide access and other restrictions.
An emerging technology, radio frequency identification (RFID) provides tracking information without requiring direct contact with the object being tracked. This is an important feature for applications like shoplifting deterrence, vehicle identification, and animal or person monitoring.
RFID systems utilize a transceiver to send digitally encoded information through an antenna via radio waves. The transceiver may be capable of both reading from and writing to RFID tags, also called transponders, which are placed on the objects being tracked. These tags may contain a battery and be able to actively send signals to the transceiver, or they may be passively controlled by the transceiver. The transceiver is linked to a computer system that interprets and manages the data being sent and received.
Many security applications use biometrics or voice recognition as a means to positively identify individuals. The most common of these is fingerprinting. While most fingerprints are still processed through the traditional manual ink method, automated fingerprint storage and identification is being used more frequently in law enforcement and other security operations, such as restricting building access and identifying account holders at banks.
Biometrics encompasses a variety of other personal identification methods, all used mostly for security. Retinal and iris scanning techniques involve identifying unique patterns in each human eye. The data describing eye features can be stored using a relatively small amount of memory, and this method may be more definitive than even fingerprinting. Another advantage is that eye scanning systems don't require direct contact with the individual, since the eyes can be scanned by a camera in the distance, and thus don't require any effort or cooperation from the human subject. Another biometric method is hand geometry identification. This method scans the shape of the back and side of a person's hand, and compares the scanned image to its database of stored shapes. This is an older technology that is still used to control access to buildings or other privileges. Individual face and even smell recognition are other biometric tools available.
Finally, voice recognition can be used for personal identification or for general data capturing. Voice-based security systems store a record of speech traits and patterns, which are difficult to counterfeit. A faster-growing use is word recognition for data entry. At its most basic, word recognition is used to identify discrete words, letters, or numbers, such as in an automated telephone menu system that allows users to say the option number of the menu item they wish to choose. More powerful software actually processes continuous speech and turns it into electronic data. This capability is now appearing in many consumer software applications, and has many uses in commercial applications.