Downloaded From : EasyEngineering.

net

Chapter 12

Automatic Identification
Ea
syE and Data Capture
ngi
CHAPTER CONTENTS nee
rin
12.1 Overview of Automatic Identification Methods
12.2 Bar Code Technology

g
12.2.1 Linear (One-Dimensional) Bar Codes
12.2.2 Two-Dimensional Bar Codes
12.3 Radio Frequency Identification
12.4 Other AIDC Technologies
12.4.1 Magnetic Stripes .ne
12.4.2 Optical Character Recognition
12.4.3 Machine Vision
t
Automatic identification and data capture (AIDC) refers to technologies that provide
direct entry of data into a computer or other microprocessor-controlled system with-
out using a keyboard. Many of these technologies require no human involvement in
the data capture and entry process. Automatic identification systems are being used
increasingly to collect data in material handling and manufacturing applications. In
material handling, the applications include shipping and receiving, storage, sortation,
order picking, and kitting of parts for assembly. In manufacturing, the applications
include monitoring the status of order processing, work-in-process, machine utiliza-
tion, worker attendance, and other measures of factory operations and performance.
Of course, AIDC has many important applications outside the factory, including retail
sales and inventory control, warehousing and distribution center operations, mail

337

Downloaded From : EasyEngineering.net

 Downloaded From : EasyEngineering.net

338 Chap. 12 / Automatic Identification and Data Capture

and parcel handling, patient identification in hospitals, check processing in banks,

and security systems. This chapter emphasizes material handling and manufacturing
applications.
The alternative to automatic data capture is manual collection and entry of
data. This typically involves a worker recording the data on paper and later entering
them into the computer by means of a keyboard. There are several drawbacks to this
method:

1. Errors occur in both data collection and keyboard entry of the data when it is
accomplished manually. The average error rate of manual keyboard entry is one
error per 300 characters.
2. Time factor. Manual methods are inherently more time consuming than automated
methods. Also, when manual methods are used, there is a time delay between when

Ea
the activities and events occur and when the data on status are entered into the
computer.
3. Labor cost. The full attention of human workers is required in manual data collec-

syE tion and entry, with the associated labor cost.

These drawbacks are virtually eliminated when automatic identification and data cap-

ngi
ture are used. With AIDC, the data on activities, events, and conditions are acquired at
the location and time of their occurrence and entered into the computer immediately or
shortly thereafter.

nee
Automatic data capture is often associated with the material handling indus-
try. The AIDC industry trade association, the Automatic Identification Manufacturers
Association (AIM), started as an affiliate of the Material Handling Institute, Inc. Many of
the applications of this technology relate to material handling. But automatic identifica-

rin
tion and data capture has also become important in shop floor control in manufacturing
plants (Section 25.4).

12.1 OVERVIEW OF AUTOMATIC IDENTIFICATION METHODS

g .ne
Nearly all of the automatic identification technologies consist of three principal compo-
nents, which also comprise the sequential steps in AIDC [8]:

t
1. Data encoder. A code is a set of symbols or signals that usually represent alpha-
numeric characters. When data are encoded, the characters are translated into a
machine-readable code. (For most AIDC techniques, the encoded data are not
readable by humans.) A label or tag containing the encoded data is attached to the
item that is to be identified.
2. Machine reader or scanner. This device reads the encoded data, converting them to
alternative form, usually an electrical analog signal.
3. Data decoder. This component transforms the electrical signal into digital data and
finally back into the original alphanumeric characters.

Many different technologies are used to implement automated identification and

data collection. Within the category of bar codes alone (currently the leading AIDC

Downloaded From : EasyEngineering.net

 Downloaded From : EasyEngineering.net

Sec. 12.1 / Overview of Automatic Identification Methods 339

technology), more than 250 different bar code schemes have been devised. AIDC tech-
nologies can be divided into the following six categories [18]:

1. Optical. Most of these technologies use high-contrast graphical symbols that can be
interpreted by an optical scanner. They include linear (one-dimensional) and two-
dimensional bar codes, optical character recognition, and machine vision.
2. Electromagnetic. The important AIDC technology in this group is radio frequency
identification (RFID), which uses a small electronic tag capable of holding more
data than a bar code. Its applications are gaining on bar codes due to several
mandates from companies like Walmart and from the U.S. Department of
Defense.
3. Magnetic. These technologies encode data magnetically, similar to recording tape.
The two important techniques in this category are (a) magnetic stripe, widely used

Ea in plastic credit cards and bank access cards, and (b) magnetic ink character recog-
nition, widely used in the banking industry for check processing.

syE 4. Smart card. This term refers to small plastic cards (the size of a credit card) imbedded
with microchips capable of containing large amounts of information. Other terms
used for this technology include chip card and integrated circuit card.
5. Touch techniques. These include touch screens and button memory.

ngi
6. Biometric. These technologies are utilized to identify humans or to interpret vocal
commands of humans. They include voice recognition, fingerprint analysis, and reti-
nal eye scans.

nee
The most widely used AIDC technologies in production and distribution are bar
codes and radio frequency methods. The common applications of AIDC technologies are

rin
(1) receiving, (2) shipping, (3) order picking, (4) finished goods storage, (5) manufacturing
processing, (6) work-in-process storage, (7) assembly, and (8) sortation. Some of the iden-

g
tification applications require workers to be involved in the data collection procedure,

.ne
usually to operate the identification equipment in the application. These techniques are
therefore semiautomated rather than automated methods. Other applications accomplish
the identification with no human participation. The same basic sensor technologies may

t
be used in both cases. For example, certain types of bar code readers are operated by hu-
mans, whereas other types operate automatically.
As indicated in the chapter introduction, there are good reasons for using auto-
matic identification and data capture techniques: (1) data accuracy, (2) timeliness, and
(3) labor reduction. First and foremost, the accuracy of the data collected is improved
with AIDC, in many cases by a significant margin. The error rate in bar code technol-
ogy is approximately 10,000 times lower than in manual keyboard data entry. The error
rates of most of the other technologies are not as low as for bar codes but are still better
than manual methods. The second reason for using automatic identification techniques
is to reduce the time required by human workers to make the data entry. The speed of
data entry for handwritten documents is approximately 5–7 characters/sec and it is 10–15
characters/sec (at best) for keyboard entry [16]. Automatic identification methods are
capable of reading hundreds of characters per second. The time savings from using auto-
matic identification techniques can mean substantial labor cost benefits for large plants
with many workers.

Downloaded From : EasyEngineering.net

 Downloaded From : EasyEngineering.net

340 Chap. 12 / Automatic Identification and Data Capture

TABLE 12.1 Comparison of AIDC Techniques and Manual Keyboard Data Entry

Time to Error Equipment

Technique Enter* Rate** Cost Advantages/(Disadvantages)
Manual entry Slow High Low Low initial cost
(Requires human operator)
Bar codes: 1-D Medium Low Low High speed
Good flexibility
(Low data density)
Bar codes: 2-D Medium Low High High speed
High data density
Radio frequency Fast Low High Label need not be visible to reader
Read-write capability available
High data density
(Expensive labeling)

Ea
Magnetic stripe Medium Low Medium Much data can be encoded
Data can be changed
(Vulnerable to magnetic fields)

syE
OCR (optical character
recognition)
Medium Medium Medium
(Contact required for reading)
Can be read by humans
(Low data density)
(High error rate)
Machine vision Fast
Source: Based on data from Palmer [14].
ngi *** Very high High speed

nee
*Time to enter data is based on a 20-character field. All techniques except machine vision use a human worker either to enter the data (manual
entry) or to operate the AIDC equipment (bar codes, RFID, magnetic stripe, OCR). Key: Slow = 5 - 10 sec, Medium = 2 - 5 sec, Fast = 6 2 sec
**Substitution error rate (SER).
***Application dependent.

rin
Although the error rate in automatic identification and data collection technologies

g
is much lower than for manual data collection and entry, errors do occur in AIDC. The

.ne
industry has adopted two parameters to measure the errors:

1. First read rate (FRR). This is the probability of a successful (correct) reading by the

t
scanner in its initial attempt.
2. Substitution error rate (SER). This is the probability or frequency with which the
scanner incorrectly reads the encoded character as some other character. In a given
set of encoded data containing n characters, the expected number of errors = SER
multiplied by n.

Obviously, it is desirable for the AIDC system to possess a high first read rate and a low
substitution error rate. A subjective comparison of substitution error rates for several
AIDC technologies is presented in Table 12.1.

12.2 BAR CODE TECHNOLOGY

As mentioned previously, bar codes divide into two basic types: (1) linear, in which the
encoded data are read using a linear sweep of the scanner, and (2) two-dimensional, in
which the encoded data must be read in both directions.

Downloaded From : EasyEngineering.net

 Downloaded From : EasyEngineering.net

Sec. 12.2 / Bar Code Technology 341

3 22170 22024 9

(a) (b)

Figure 12.1 Two forms of linear bar codes are (a) width-modulated, exempli-
fied here by the Universal Product Code, and (b) height-modulated, exemplified
here by Postnet, used by the U.S. Postal Service.

Ea 12.2.1 Linear (One-Dimensional) Bar Codes

syE Linear bar codes are the most widely used automatic identification and data capture
technique. There are actually two forms of linear bar code symbologies, illustrated in
Figure 12.1: (a) width-modulated, in which the symbol consists of bars and spaces of
varying width; and (b) height-modulated, in which the symbol consists of evenly spaced

ngi
bars of varying height. The only significant application of the height-modulated bar code
symbologies is in the U.S. Postal Service for ZIP code identification, so the discussion
here focuses on the width-modulated bar codes, which are used widely in retailing and
manufacturing.

nee
In linear width-modulated bar code technology, the symbol consists of a sequence of
wide and narrow colored bars separated by wide and narrow spaces (the colored bars are

rin
usually black and the spaces are white for high contrast). The pattern of bars and spaces is
coded to represent numeric or alphanumeric characters. Palmer [14] uses the interesting
analogy that bar codes might be thought of as a printed version of the Morse code, where

g
narrow bands represent dots and wide bands represent dashes. Using this scheme, the bar

.ne
code for the familiar SOS distress signal would be as shown in Figure 12.2. Bar codes do
not follow Morse code, however; the difficulties with a “Morse” bar code symbology are
(1) only the dark bars are used, thus increasing the length of the coded symbol, and (2) the

t
number of bars making up the alphanumeric characters differs, making decoding more
difficult [14].
Bar code readers interpret the code by scanning and decoding the sequence of bars.
The reader consists of the scanner and decoder. The scanner emits a beam of light that
is swept past the bar code (either manually or automatically) and senses light reflections
to distinguish between the bars and spaces. The light reflections are sensed by a photo-
detector, which converts the spaces into an electrical signal and the bars into absence of
an electrical signal. The width of the bars and spaces is indicated by the duration of the

Figure 12.2 The SOS distress signal in “Morse” bar codes.

Downloaded From : EasyEngineering.net

 Downloaded From : EasyEngineering.net

342 Chap. 12 / Automatic Identification and Data Capture

Sweep of
light beam

(a)

(b)

Figure 12.3 Conversion of bar code into a pulse train of electrical

signals, (a) bar code and (b) corresponding electrical signal.

Ea
corresponding signals. The procedure is depicted in Figure 12.3. The decoder analyzes

syE
the pulse train to validate and interpret the corresponding data.
Certainly a major reason for the acceptance of bar codes is their widespread use
in grocery markets and other retail stores. In 1973, the grocery industry adopted the
Universal Product Code (UPC) as its standard for item identification. This is a 12-digit

ngi
bar code that uses six digits to identify the manufacturer and five digits to identify the
product. The final digit is a check character. The U.S. Department of Defense provided
another major endorsement in 1982 by adopting a bar code standard (Code 39) that must

nee
be applied by vendors on product cartons supplied to the various agencies of DOD. The
UPC is a numerical code (0–9), while Code 39 provides the full set of alphanumeric char-
acters plus other symbols (44 characters in all). These two linear bar codes and several
others are compared in Table 12.2.

rin
g
TABLE 12.2 Some Widely Used Linear Bar Codes
Bar Code
Codabar
Description
Only 16 characters: 0–9,
$, :, /, ., +, -
Applications

.ne
Used in libraries, blood banks, and
some parcel freight applications
UPC*

Code 39
Numeric only,
length = 12 digits

Alphanumeric (see text for

description)
stores t
Widely used in the United States and
Canada, in grocery and other retail

Adopted by Department of Defense,

automotive, and other manufacturing
industries
Postnet Numeric only** U.S. Postal Service code for ZIP code
numbers
Code 128 Alphanumeric, but higher Substitutes in some Code 39
density applications
Code 93 Similar to Code 39 but Same applications as Code 39
higher density
Sources: Nelson [13], Palmer [14].
*UPC = Universal Product Code, adopted by the grocery industry in 1973 and based on a symbol developed by
IBM Corporation in early grocery tests. A similar standard bar code system was developed in Europe, called the
European Article Numbering system (EAN), in 1978.
**This is the only height-modulated bar code in the table. All others are width-modulated.

Downloaded From : EasyEngineering.net

 Downloaded From : EasyEngineering.net

Sec. 12.2 / Bar Code Technology 343

The Bar Code Symbol. The bar code standard adopted by the automotive in-
dustry, the Department of Defense, the General Services Administration, and many
other manufacturing industries is Code 39, also known as AIM USD–2 (Automatic
Identification Manufacturers Uniform Symbol Description-2). Code 39 uses a series
of wide and narrow elements (bars and spaces) to represent alphanumeric and other
characters. The wide elements are equivalent to a binary value of one and the narrow
elements are equal to zero. The width of the wide bars and spaces is between two and
three times the width of the narrow bars and spaces. Whatever the wide-to-narrow
ratio, the width must be uniform throughout the code for the reader to be able to con-
sistently interpret the resulting pulse train. Figure 12.4 presents the character structure
for USD–2.
The reason for the name Code 39 is that nine elements (bars and spaces) are used
in each character and three of the elements are wide. The placement of the wide spaces

Ea
and bars in the code uniquely designates the character. Each code begins and ends with
either a wide or narrow bar. The code is sometimes referred to as code three-of-nine.
In addition to the character set in the bar code, there must also be a so-called “quiet

syE zone” both preceding and following the bar code, in which there is no printing that
might confuse the decoder.

ngi
Bar Code Readers. Bar code readers come in a variety of configurations; some
require a human to operate them and others are stand-alone automatic units. They are
usually classified as contact or noncontact readers. Contact bar code readers are hand-

nee
held wands or light pens operated by moving the tip of the wand quickly past the bar code
on the object or document. The wand tip must be in contact with the bar code surface
or in very close proximity during the reading procedure. In a factory data collection ap-
plication, they are usually part of a keyboard entry terminal. The terminal is sometimes

rin
referred to as a stationary terminal in the sense that it is placed in a fixed location in the
shop. When a transaction is entered in the factory, the data are usually communicated

g
to the computer system immediately. In addition to their use in factory data collection

.ne
systems, stationary contact bar code readers are widely used in retail stores to enter the
item in a sales transaction.
Contact bar code readers are also available as portable units that can be carried

t
around the factory or warehouse by a worker. They are battery-powered and include
a solid-state memory device capable of storing data acquired during operation. The
data can be transferred to the computer system subsequently. Portable bar code read-
ers often include a keypad that can be used by the operator to input data that cannot
be entered via bar code. These portable units are used for order picking in a ware-
house and similar applications that require a worker to move significant distances in a
building.
Noncontact bar code readers focus a light beam on the bar code, and a photo-
detector reads the reflected signal to interpret the code. The reader probe is located
a certain distance from the bar code (several inches to several feet) during the read
procedure. Noncontact readers are classified as fixed beam and moving beam scan-
ners. Fixed beam readers are stationary units that use a fixed beam of light. They are
usually mounted beside a conveyor and depend on the movement of the bar code past
the light beam for their operation. Applications of fixed beam bar code readers are
typically in warehousing and material handling operations where large quantities of
materials must be identified as they flow past the scanner on conveyors. Fixed beam

Downloaded From : EasyEngineering.net

 Downloaded From : EasyEngineering.net

344 Chap. 12 / Automatic Identification and Data Capture

Char. Bar pattern 9 bits Char. Bar pattern 9 bits

1 100100001 K 100000011

2 001100001 L 001000011

3 101100000 M 101000010

4 000110001 N 000010011

5 100110000 O 100010010

6 001110000 P 001010010

Ea
7 000100101 Q 000000111

8 100100100 R 100000110

syE 9

0
001100100

000110100
S

T
001000110

000010110

Bngi 100001001

001001001
U

V
110000001

011000001

D
nee
101001000

000011001
W

X
111000000

010010001

E 100011000 Y
rin 110010000

H
001011000

000001101

100001100
Z

.
g 011010000

.ne
010000101

110000100

J
001001100

000011100
space

*
011000100

010010100
t
*Denotes a start/stop code that must be placed at the beginning and
end of every bar code message.

Figure 12.4 Character set in USD-2 bar code, a subset of

Code 39 [6].

scanners in these kinds of operations represent some of the first applications of bar
codes in industry.
Moving beam scanners use a highly focused beam of light, often a laser, actuated
by a rotating mirror to traverse an angular sweep in search of the bar code on the object.
A scan is defined as a single sweep of the light beam through the angular path. The high

Downloaded From : EasyEngineering.net

 Downloaded From : EasyEngineering.net

Sec. 12.2 / Bar Code Technology 345

Carton
Bar code

Conveyor
Moving beam

Scanner

Figure 12.5 Stationary moving beam bar code scanner located along a
moving conveyor.

Ea rotational speed of the mirror allows for very high scan rates—up to 1,440 scans/sec [1].
This means that many scans of a single bar code can be made during a typical reading pro-

syE cedure, thus permitting verification of the reading. Moving beam scanners can be either
stationary or portable units. Stationary scanners are located in a fixed position to read bar
codes on objects as they move past on a conveyor or other material handling equipment.

ngi
They are used in warehouses and distribution centers to automate the product identifica-
tion and sortation operations. A typical setup using a stationary scanner is illustrated in
Figure 12.5. Portable scanners are handheld devices that the user points at the bar code

nee
like a pistol. The vast majority of bar code scanners used in factories and warehouses are
of this type.

Bar Code Printers. In many bar code applications, the labels are printed in

rin
medium-to-large quantities for product packages and the cartons used to ship the pack-
aged products. These preprinted bar codes are usually produced off-site by companies

g
specializing in these operations. The labels are printed in either identical or sequenced
symbols. Printing technologies include traditional techniques such as letterpress, offset
lithography, and flexographic printing.

.ne
Bar codes can also be printed on-site by methods in which the process is controlled

t
by microprocessor to achieve individualized printing of the bar coded document or item
label. These applications tend to require multiple printers distributed at locations where
they are needed. The printing technologies used in these applications include ink-jet,
laser printing, and laser etching [7], [9], [14].
Examples of applications of these individualized bar code printing methods include
keyboard entry of data for inclusion in the bar code of each item that is labeled, unique
identification of production lots for pharmaceutical products, and preparation of route
sheets and other documents included in a shop packet traveling with a production order.
Production workers use bar code readers to indicate order number and completion of
each step in the operation sequence.

12.2.2 Two-Dimensional Bar Codes

The first two-dimensional (2-D) bar code was introduced in 1987. Since then, more than a
dozen 2-D symbol schemes have been developed, and the number is expected to increase.
The advantage of 2-D codes is their capacity to store much greater amounts of data at

Downloaded From : EasyEngineering.net

 Downloaded From : EasyEngineering.net

346 Chap. 12 / Automatic Identification and Data Capture

higher area densities. Their disadvantage is that special scanning equipment is required
to read the codes, and the equipment is more expensive than scanners used for conven-
tional bar codes. Two-dimensional symbologies divide into two basic types: (1) stacked
bar codes and (2) matrix symbologies.

Stacked Bar Codes. The first 2-D bar code to be introduced was a stacked sym-
bology. It was developed in an effort to reduce the area required for a conventional
bar code. But its real advantage is that it can contain significantly greater amounts of
data. A stacked bar code consists of multiple rows of conventional linear bar codes
stacked on top of each other. Several stacking schemes have been devised over the
years, nearly all of which allow for multiple rows and variations in the numbers of
encoded characters possible. An example of a 2-D stacked bar code is illustrated in
Figure 12.6. The data density of stacked bar codes is typically five to seven times that

Ea
of the linear bar code 39.
The encoded data in a stacked bar code are decoded using laser-type scanners that
read the lines sequentially. The technical problems encountered in reading a stacked

syE
bar code include (1) keeping track of the different rows during scanning, (2) dealing
with scanning swaths that cross between rows, and (3) detecting and correcting local-
ized errors [14]. As in linear bar codes, printing defects in the 2-D bar codes are also a

ngi
problem.

Matrix Symbologies. A matrix symbology consists of 2-D patterns of data cells

nee
that are usually square and are colored dark (usually black) or white. The 2-D matrix
symbologies were introduced around 1990. Their advantage over stacked bar codes
is their capability to contain more data. They also have the potential for higher data
densities—up to 30 times more dense than Code 39. Their disadvantage compared to

rin
stacked bar codes is that they are more complicated, which requires more sophisticated
printing and reading equipment. The symbols must be produced (during printing) and

g
interpreted (during reading) both horizontally and vertically; therefore, they are some-
times referred to as area symbologies. An example of a 2-D matrix code is illustrated

.ne
t

Figure 12.6 A 2-D stacked bar code. Shown is an example of a

PDF417 symbol.

Downloaded From : EasyEngineering.net

 Downloaded From : EasyEngineering.net

Sec. 12.3 / Radio Frequency Identification 347

Ea Figure 12.7 A 2-D matrix bar code. Shown is

syE an example of the Data Matrix symbol.

ngi
in Figure 12.7. Reading a Data Matrix code requires a machine vision system specially
programmed to interpret the code.

nee
Applications of the matrix symbologies are found in part and product identification
during manufacturing and assembly. These kinds of applications are expected to grow
as computer-integrated manufacturing becomes more pervasive throughout industry.
The semiconductor industry has adopted Data Matrix ECC200 (a variation of the Data

rin
Matrix code shown in Figure 12.7) as its standard for marking and identifying wafers and
other electronic components [12].

12.3 RADIO FREQUENCY IDENTIFICATION g .ne

t
Radio frequency identification technology (RFID) represents the biggest challenge to
the dominance of bar codes. Companies including Walmart, Target, and Metro AG
(in  Germany), as well as the U.S. Department of Defense, have mandated that their
suppliers use RFID on incoming materials. In fact the Department of Defense requires
a combination of Data Matrix and RFID on all “mission-critical” parts, assemblies, and
equipment. These requirements have provided a significant impetus for the implementa-
tion of RFID in industry. According to a study of Walmart cited in [17], “RFID stores
are 63 percent more effective in replenishing out-of-stock items than traditional stores.”
In radio frequency identification, an identification tag or label containing elec-
tronically encoded data is attached to the subject item, which can be a part, product, or
container (e.g., carton, tote pan, pallet). The identification tag consists of an integrated
circuit chip and a small antenna, as pictured in Figure 12.8. These components are usu-
ally enclosed in a protective plastic container or are imbedded in an adhesive-backed
label that is attached to item. The tag is designed to satisfy the Electronic Product
Code (EPC) standard, which is the RFID counterpart to the Universal Product Code
(UPC) used in bar codes. The tag communicates the encoded data by RF to a reader

Downloaded From : EasyEngineering.net

 Downloaded From : EasyEngineering.net

348 Chap. 12 / Automatic Identification and Data Capture

lntegrated circuit chip Antenna

Ea Adhesive-backed label

syE Figure 12.8 RFID label. Approximate size is 20 mm by

30 mm (0.8 in by 1.2 in).

ngi
or interrogator as the item is brought into the reader’s proximity. The reader can be

nee
portable or stationary. It decodes and confirms the RF signal before transmitting the
associated data to a collection computer.
Although the RF signals are similar to those used in wireless radio and television
transmission, there are differences in how RF technology is used in product identifica-

rin
tion. One difference is that the communication is in two directions rather than in one
direction as in commercial radio and TV. The identification tag is a transponder, a de-

g
vice that emits a signal of its own when it receives a signal from an external source. To

.ne
activate it, the reader transmits a low-level RF magnetic field that serves as the power
source for the transponder when they are near each other. Another difference between
RFID and commercial radio and TV is that the signal power is substantially lower in

t
RFID applications (milliwatts to several watts), and the communication distances usually
range between several millimeters and several meters. Finally, there are differences in the
allowable frequencies that can be used for RFID applications versus radio, TV, and other
commercial and military users.
RF identification tags are available in two general types: (1) passive and (2) active.
Passive tags have no internal power source; they derive their electrical power for trans-
mitting a signal from radio waves generated by the reader when in close proximity. Active
tags include their own battery power packs. Passive tags are smaller, less expensive, lon-
ger lasting, and have a shorter radio communication range. Active tags generally possess
a larger memory capacity and a longer communication range (typically 10 m and more).
Applications of active tags tend to be associated with higher value items due to the higher
cost per tag.
One of the initial uses of RFID was in Britain in World War II to distinguish between
enemy and allied airplanes flying across the English Channel. Commercial and military
aircraft still use transponders for identification purposes. Other RFID applications use
tags in a variety of different forms, such as credit-card-sized plastic labels for product

Downloaded From : EasyEngineering.net

 Downloaded From : EasyEngineering.net

Sec. 12.4 / Other AIDC Technologies 349

TABLE 12.3 Bar Codes versus Radio Frequency Identification

Comparison Bar Codes RFID

Technology Optical Radio frequency
Read-write capability Read only Read-write available
Storage capacity 14–16 digits 96–256 digits
(linear bar codes)
Line-of-sight reading Required Not required
Reusability One-time use Reusable
Cost per label Very low About 10 times the cost of
bar code for passive tag
Durability Susceptible to dirt and  More durable in plant
scratches environment
Source: Based mostly on Weber [17].

Ea identification and very small glass capsules injected into wild animals for tracking and re-

syE search purposes. The principal applications of RFID in industry (in approximate descend-
ing order of frequency) are (1) inventory management, (2) supply chain management,
(3) tracking systems, (4) warehouse control, (5) location identification, and (6) work-in-

ngi
process tracking [17].
Identification tags in RFID have traditionally been read-only devices that contain
up to 20 characters of data identifying the item and representing other information that

nee
is to be communicated. Advances in the technology have provided much higher data
storage capacity and the ability to change the data in the tag (read/write tags). This has
opened opportunities for incorporating much more status and historical information into
the automatic identification tag rather than using a central database. Table 12.3 compares
the two major AIDC technologies, bar codes and RFID.

rin
Advantages of RFID include the following: (1) identification does not depend on

g
physical contact or direct line of sight observation by the reader, (2) much more data can
be contained in the identification tag than with most AIDC technologies, and (3) data

.ne
in the read/write tags can be altered for historical usage purposes or reuse of the tag.
The disadvantage of RFID is that the labels and hardware are more expensive than for

t
most other AIDC technologies. For this reason, RFID systems have traditionally been
appropriate only for data collection situations in which environmental factors preclude
the use of optical techniques such as bar codes to identify products in manufacturing pro-
cesses that would obscure any optically coded data (e.g., spray painting). The applications
are now expanding beyond these limits due to the mandates set forth by Walmart, the
Department of Defense, and others.
In addition to RF identification, radio frequencies are also widely used to augment
bar code and other AIDC techniques by providing the communication link between re-
mote bar code readers and some central terminal. This latter application is called radio
frequency data communication (RFDC), as distinguished from RFID.

12.4 OTHER AIDC TECHNOLOGIES

The other automated identification and data collection techniques are either used in spe-
cial applications in factory operations, or they are widely applied outside the factory.

Downloaded From : EasyEngineering.net

 Downloaded From : EasyEngineering.net

350 Chap. 12 / Automatic Identification and Data Capture

12.4.1 Magnetic Stripes

Magnetic stripes attached to a product or container are sometimes used for item iden-
tification in factory and warehouse applications. A magnetic stripe is a thin plastic film
containing small magnetic particles whose pole orientations can be used to encode bits
of data into the film. The film can be encased in or attached to a plastic card or paper
ticket for automatic identification. These are the same kinds of magnetic stripes used
to encode data onto plastic credit cards and bank access cards. Two advantages of mag-
netic stripes are their large data storage capacity and the ability to alter the data con-
tained in them. Although they are widely used in the financial community, their use
seems to be declining in shop floor control applications for the following reasons: (1)
the magnetic stripe must be in contact with the scanning equipment for reading to be ac-
complished, (2) there are no convenient shop floor encoding methods to write data into

Ea
the stripe, and (3) the magnetic stripe labels are more expensive than bar code labels.

12.4.2 Optical Character Recognition

syE
Optical character recognition (OCR) is the use of specially designed alphanumeric
characters that are machine readable by an optical reading device. Optical character

ngi
recognition is a 2-D symbology, and scanning involves interpretation of both the ver-
tical and horizontal features of each character during decoding. Accordingly, when
manually operated scanners are used, a certain level of skill is required by the human
operator, and first read rates are relatively low (often less than 50% [14]). The substan-

humans as well as by machines.

nee
tial benefit of OCR technology is that the characters and associated text can be read by

As an interesting historical note, OCR was selected as the standard automatic iden-

rin
tification technology by the National Retail Merchants Association (NRMA) shortly
after the UPC bar code was adopted by the grocery industry. Many retail establishments

g
made the investment in OCR equipment at that time. However, the problems with the
technology became apparent by the mid-1980s [14]: (1) low first read rate and high sub-

.ne
stitution error rate when handheld scanners were used, (2) lack of an omnidirectional
scanner for automatic checkout, and (3) widespread and growing adoption of bar code
technology. NRMA was subsequently forced to revise its recommended standard from
OCR technology to bar codes.

t
For factory and warehouse applications, the list of disadvantages includes (1) the re-
quirement for near-contact scanning, (2) lower scanning rates, and (3) higher error rates
compared to bar code scanning.

12.4.3 Machine Vision

The principal application of machine vision is for automated inspection tasks (Section 22.5).
For AIDC applications, machine vision systems are used to read 2-D matrix symbols, such
as Data Matrix (Figure 12.7), and they can also be used for stacked bar codes, such as PDF-
417 (Figure 12.6) [11]. Applications of machine vision also include other types of automatic
identification, and these applications may grow in number as the technology advances.
For example, machine vision systems are capable of distinguishing among a variety of prod-
ucts moving down a conveyor so that the products can be sorted. The recognition task
is accomplished without using special identification codes on the products and is instead
based on the inherent geometric features of the object.

Downloaded From : EasyEngineering.net

 Downloaded From : EasyEngineering.net

Review Questions 351

REFERENCES

[1] Accu-Sort Systems, Inc., Bar Code Technology—Present State, Future Promise, 2nd ed.,
Telford, PA (no date).
[2] “AIDC Technologies—Who Uses Them and Why,” Modern Materials Handling, March
1993, pp. 12–13.
[3] Agapakis, J., and A. Stuebler, “Data Matrix and RFID—Partnership in Productivity,”
Assembly, October 2006, pp. 56–59.
[4] Allais, D. C., Bar Code Symbology, Intermec Corporation, 1984.
[5] Attaran, M., “RFID Pays Off,” Industrial Engineer, September 2006, pp. 46–50.
[6] Automatic Identification Manufacturers, Automatic Identification Manufacturers Manual,
Pittsburgh, PA.
[7] “Bar Codes Move into the Next Dimension,” Modern Materials Handling/AIDC News &

Ea
Solutions, June 1998, p. A11.
[8] Cohen, J., Automatic Identification and Data Collection Systems, McGraw-Hill Book
Company Europe, Berkshire, UK, 1994.

syE [9]

[10]
Forcino, H., “Bar Code Revolution Conquers Manufacturing,” Managing Automation, July
1998, pp. 59–61.
Kinsella, B., “Delivering the Goods,” Industrial Engineer, March 2005, pp. 24–30.
[11]

[12]
pp. 73–77.
ngi
Moore, B., “New Scanners for 2D Symbols,” Material Handling Engineering, March 1998,

Navas, D., “Vertical Industry Overview: Electronics ‘98,” ID Systems, February 1998,

nee
pp. 16–26.
[13] Nelson, B., Punched Cards to Bar Codes, Helmers Publishing, Inc., Peterborough, NH, 1997.
[14] Palmer, R. C., The Bar Code Book, 5th ed., Helmers Publishing, Inc., Peterborough, NH,
2007.
[15]
December 2003, pp. 24–30.
rin
“RFID: Wal-Mart Has Spoken. Will You Comply?” Material Handling Management,

[16]

[17]
[18]
Industrial Engineering, November 1985, pp. 55–59.
g
Soltis, D. J., “Automatic Identification System: Strengths, Weaknesses, and Future Trends,”

Weber, A., “RFID on the Line,” Assembly, January 2006, pp. 78–92.
www.aimusa.org/techinfo/aidc.html
.ne
REVIEW QUESTIONS
[19]
[20]
www.wikipedia.org/wiki/Barcode
www.wikipedia.org/wiki/Radio_frequency_identification
t
12.1 What is automatic identification and data capture?
12.2 What are the drawbacks of manual collection and entry of data?
12.3 What are the three principal components in automatic identification technologies?
12.4 Name four of the six categories of AIDC technologies that are identified in the text.
12.5 Name five common applications of AIDC technologies in production and distribution.
12.6 There are two forms of linear bar codes. Name them, and indicate what the difference is.
12.7 What was the major industry to first use the Universal Product Code (UPC)?
12.8 What are the two basic types of two-dimensional bar codes?
12.9 What does RFID stand for?

Downloaded From : EasyEngineering.net

 Downloaded From : EasyEngineering.net

352 Chap. 12 / Automatic Identification and Data Capture

12.10 What is a transponder in RFID?

12.11 What is the difference between a passive tag and an active tag?
12.12 What are the relative advantages of RFID over bar codes?
12.13 What are the relative advantages of bar codes over RFID?
12.14 What are the reasons why magnetic stripes are not widely used in factory floor
operations?
12.15 What is the advantage of optical character recognition technology over bar code
technology?
12.16 What is the principal application of machine vision in industry?

Ea
syE
ngi
nee
rin
g .ne
t

Downloaded From : EasyEngineering.net