HomeAbout UsBrowse This IssueBack IssuesNews DispatchesSubscribing to Army LogisticianWriting for Army LogisticianContact UsLinks































Joint Asset Visibility:
Why So Hard?
Capturing Information

In the second of his articles on joint asset visibility, the author discusses the processes and systems used for capturing data for asset visibility.

Department of Defense (DOD) stakeholders are interested in obtaining a lot of information about items in transit or in storage. Information about items in storage is usually reported whenever on-hand balances change, such as when additional items are received or issued. However, this information is usually updated on a daily basis as it is passed electronically through the supply chain—from the unit level to battalion, brigade, division, theater, and onwards to the strategic-level—using wide-area networks such as the Joint Operations Planning and Execution System (JOPES), the Global Transportation Network (GTN), or the Defense Automatic Addressing System (DAAS).

In an ideal world, the worldwide DOD item balances would be updated automatically any time ownership or location of an item changed. In reality, however, this is not the case; the limitations of telecommunications and the time needed for some computers to process information prevent real-time displays.

For example, consider high-mobility multipurpose wheeled vehicle (HMMWV) tires. Let’s say that, on a particular day, the various units and DOD agencies have a worldwide total of 100,000 tires stored or in transit at 5,000 different sites. How could logisticians keep track of the on-hand balances of all of the HMMWV tires over time as tires were purchased, issued, condemned, and transferred at the various locations? After all, not all of these sites would be connected to the World Wide Web. (For instance, a large portion of a deployed ground force may be powered by generators or may have no electrical power source at all. Without access to an electrical grid, it is much more difficult to connect to the Web.) For units not connected to the Web, logistics information is still passed along the supply chain echelons, but this is done in batch mode, not real time. A unit might have to deliver a disc to the automated information system of its supply support activity (SSA), or perhaps the SSA would have to send its asset visibility information via satellite to the automated information system at theater level. Depending on the unit’s location, the method of transmission, and the extent of information to be passed, the use of direct links to satellites can be an expensive, manpower-intensive proposition. Moreover, some legacy automated information systems need time to process account balances, putting all other computer processes on hold while preparing to transmit data to a higher or lower echelon. Instead of being passed and processed in real-time, the information is passed and processed in a batch. A batch of information may be passed at various time intervals, such as hourly, twice a day, four times a day, daily, weekly, or monthly. However, for the most part, stakeholders within the DOD global supply chain are looking for daily asset visibility reports for supplies and equipment in storage or on hand at the unit level. For instance, most DOD lostitics managers would be more than satisfied with daily on-hand availability updates by location of HMMWV tires (or other item).

Although DOD supply chain stakeholders can accept in-storage information that is updated daily, they want updates about items in-transit on a more frequent basis. Ideally, stakeholders want to know when an item in transit arrives and departs a transshipment point.

Items traverse a wide variety of transshipment points “from factory to foxhole.” As soon as DOD purchases an item from a manufacturer or a retailer, it begins accounting for that item. Information about the item must be captured, preferably electronically, and passed from one automated information system to another until it reaches a wide-area network, accessible by all DOD logisticians. However, different transshipment points capture different types of logistics information. Depots capture wholesale information. Airports managed by the Air Force capture air transport information. Strategic deployment and distribution centers or seaports run by the Navy capture information pertaining to surface delivery. Managers at commercially run transshipment points capture logistics information that is pertinent to their particular needs. Personnel at transshipment points must capture not only information about the item being shipped but also information about the type of conveyance being used to transport the item to the next transshipment point.

The Data Capture Challenge

Unfortunately, many transshipment points do not have the manpower, computing power, or telecommunications equipment needed to capture all of the pertinent logistics information. This is especially true when items in transit are substantially reconfigured.

Let’s look at a hypothetical scenario. To support forces deployed throughout Iraq, Defense Logistics Agency (DLA) personnel stationed at the Defense Distribution Depot Susquehanna, Pennsylvania, load 4,000 different items (with 4,000 different national stock numbers [NSNs] and 4,000 different document numbers) into one 40-foot container. The automated information system at this sophisticated transshipment point electronically captures the NSN and document number for each item within the container. This information is then passed to the wide-area networks, such as JOPES, DAAS, and GTN.

Next, the container is sent to a military or commercial port in the continental United States, where the supply and transportation data are again easily captured. This is not that difficult because the contents of the 40-foot container remain the same; only the container’s location has changed. While en route to Iraq, however, the vessel transporting the container turns out to be too big to transit the Suez Canal. The vessel’s draft also is too deep to access the overseas port of debarkation. As a result, the contents of the 40-foot container have to be offloaded at an intermediary port at the entrance of the Suez Canal and loaded into two 20-foot containers. In this case, the previous content integrity of the 40-foot container is now gone. At this new transshipment point (the intermediary port), the automated information system now must update the container information for two containers, each with 2,000 items. Whatever data processing codes were used to identify the contents of the 40-foot container must be updated to identify the contents of the two 20-foot containers. Will this commercial overseas port be able to capture this information and then pass it on to the wide-area networks? The answer depends on the port.

Actually, this problem is even more complex than it first appears. The contents of the one 40-foot container could be spread out into 5, 10, or 20 different 20-foot containers, each containing previously loaded items in addition to the contents transferred from the 40-foot container. Similarly, the contents of each of the resultant 20-foot containers could be further broken down at a subsequent transshipment point into pallets and loaded onto aircraft, railcars, and trucks.

Will transshipment points (especially the commercial ones with scant DOD representation) have the capability to capture the appropriate logistics information into automated information systems and then transmit this information to the respective wide-area networks? The answer is probably not.

Consider how much time it would take for one person to scan the two-dimensional (2D) barcodes of military shipment labels or, worse yet, manually enter content-level data into automated information systems. Even with the most rudimentary information, such as nomenclature, document number, NSN, and transportation control number data, errors naturally occur whenever human entry is required. Some studies indicate that, for every 85 keystrokes, 3 errors are made unbeknownst to the operator. And with coded information like an NSN, if the input contains a single incorrect character, the accurate code (in this case the NSN) cannot be processed by the automated information system. Capturing barcoded information is easier and less error prone, but it is still very time-consuming; it also requires the appropriate barcode readers and complementary automated information system. Radio frequency identification (RFID) offers promise, but it is not the sole solution.

Let’s consider a simpler scenario. What if the Susquehanna Depot completely stuffed a 20-foot container with a single item: concertina wire (NSN 5660–00–921–5516). In this case, there would be only one customer and only one document number. The concertina wire, destined for an SSA in Iraq, would remain intact inside a container throughout its shipment from the depot all the way to the SSA somewhere north of Baghdad. In this scenario, the automated information system at DLA could easily capture the pertinent information about the container’s contents and associate it with the identification number of the container. This information then could be readily passed to a wide-area network, where it could be viewed by interested logisticians worldwide. All they would need to know would be the document number or the container number. Shareholders would be able to track some of the movement of the container as it made its way to Iraq as long as the various transshipment points being transited had a means of capturing the container number as it arrived and departed and a means of passing this information to a wide-area network.

So Much Data!

However, let’s take a look at a more realistic scenario. Think of how much logistics data would have to be captured and passed pertaining to a vessel carrying the equivalent of 6,000 different 20-foot containers. If each container had 2,000 different items within it, 12,000,000 different items would be on board the vessel (6,000 * 2,000). How much information would we need to capture about each item? If we only needed the document number and NSN, then we would need to capture 24,000,000 different data elements (6,000 * 2,000 * 2). But if we wanted to track each item’s document number, NSN, nomenclature, unit of issue, condition code, supplementary address, required delivery date, weight, cube, and project code, then we would need to capture 120,000,000 different data elements. We also might want to track the identification numbers of all the multipacks inside the containers so that we could associate all of the NSNs within with their specific multipack identification numbers.

If each 20-foot container were to have 50 multipacks within it, then there would be 240,000 multipacks (6,000 containers * 2,000 items per container ÷ 50 items per multipack) onboard the vessel. If these multipacks were atop pallets, with each pallet holding 6 multipacks, then there would be 40,000 different pallets (240,000 multipacks ÷ 6). Each pallet should have a unique identifying number, and each pallet number should be correlated to the identifying numbers of the six multipacks on the pallet. Associating a specific item’s document number with the identifying number of its multipack, with the identifying number of its pallet, with the identifying number of its container, and with the identifying number of the vessel can be extremely complicated. After all, in the scenario just described, 1 ship is carrying 6,000 containers, 40,000 pallets, 240,000 multipacks, and 12,000,000 different items. The amount of data that would have to be loaded into automated information systems is mind-boggling. It would take a very, very long time to capture all of these data by hand or using barcode readers, especially if it were done while the ship was being loaded. (See the article, “Containerizing the Joint Force,” published in the March–April 2005 issue of Army Logistician.)

The challenge in the scenario described above is created by the enormous amount of data. The scenario assumed that the port of debarkation had sophisticated data-capture equipment and telecommunications. Let’s take a look at a different problem. What would happen if the transshipment point were located in a very austere environment, like a desert, where enablers such as electricity, computers, and telecommunications were not even remotely available? What if the transshipment point were simply a spot in the sand where the form of conveyance changes, say from a commercial truck to a military truck? How can the information about the items that were transferred be captured within automated information systems? Under our current procedures, the answer is that the information may not be captured.

Whose Job Is It To Capture the Information?

A major problem in capturing information pertaining to asset visibility (in-transit visibility in particular) is that DOD has no designated military occupational specialty or civilian equivalent specifically trained to do so. Since so many different types of transshipment points are run by so many different types of organizations, no one has been trained on how to capture the information about supplies in transit using both joint military procedures and commercial practices. No standardized automated information systems or telecommunications systems are available to capture the information and pass it to the wide-area networks (which themselves are not standardized). Many transshipment points, such as overseas seaports and railheads, may have no DOD presence at all.

Capturing Information at Transshipment Points

Just as not every transshipment point has a designated specialist, no set method has been established for capturing the required asset visibility information. Several methods of capturing data should be available to help ensure reliability. The most basic method is for clerks to capture information by manually transferring data from the shipment documents or by jotting down the logistics data shown on the item’s packaging. However, if a clerk were simply to file the information in a filing cabinet, it would not be visible to logisticians with access to the wide-area networks. It would be much better if a clerk were to enter the pertinent logistics information into an automated information system. It would be better still if he were to capture the logistics information using electronic data interchange (EDI) and automatic information technology (AIT).

Since both EDI and AIT rely on computer processing, let’s take a look at some of the rudiments of this incredibly complex field. The ability of computers and telecommunications devices to digitize information has been truly revolutionary and has been, and still is, one of the cornerstones of logistics transformation. But what do we mean when we say information has been digitized? In the most basic sense, all
computerized information can be subcomposed into what are called “bits.” The word bit began as an abbreviation for the phrase “binary digit.” The root of binary is “bi” which connotes two of something. In computer terms, binary code means either 1 or 0; it also connotes the concept of something being on or off. Just as a light switch can be turned on or off, a silicone chip can be turned on or off. Binary code, then, is a stream of some combination of the digits 0 or 1 and is used as the basis for computer processing.

The American Standard Code for Information Interchange (ASCII), a widely accepted method of encoding characters based on the English language, uses binary values. For example, the binary value of the ASCII letter M is 00 1101 and the number 2 is 11 0001. In computer processing, such binary values can depict all the letters of the alphabet (both uppercase and lowercase), all numbers, and many special characters. Binary numbers are used to compose hexadecimal numbers (with a base of 16 binary digits), which are used extensively in RFID devices.

The text and numbers included within almost every electronic document can be subcomposed into a series of 0s and 1s. As an analogy, think how a person can navigate anywhere in the world by making a series of decisions based on only two choices: go left or go right. In our digitized world, 8 bits make 1 byte, which usually represents one alphabetic character (like A, B, or C), one special character (like &, *, or ?), or two numeric digits. A kilobyte is a measure of 1,024 bytes; a megabyte is a measure of 1,048,576 bytes; and a gigabyte is a measure of 1,073,741,824 bytes. A standard $50 thumb drive can store one gigabyte.

AIT Devices

The linear barcode is the most basic of the several different types of AIT used to store a wide variety of data. The linear barcode can store 17 to 20 alphanumeric characters. It is typically used to store one key data element, such as an NSN, a document number, or a transportation control number. If all three of these numbers are on the packaging of a container, a clerk has to scan all three numbers separately to retrieve the digital information the barcodes portray. The newer and more sophisticated 2D barcodes have a greater capacity that the linear barcode; a 2D barcode can portray about 1,850 different characters and is more reliable than a linear barcode because it has several layers of data repetition as part of its design.

A clerk at a transshipment point or a storage facility can either scan the barcodes by sliding them across a fixed scanner or use a portable scanner called a portable data collection device to scan items at various locations within a warehouse or storage yard. Regardless of whether the barcode scanner is portable or fixed, it must be linked to a computer to process the digital information, although the linkages (particularly with the portable data collection device) may be wireless. The laser technology associated with barcode readers must be able to see the barcode. In other words, the barcode must be within the barcode reader’s line of sight. Humans must be involved in lining up an item’s barcode with the barcode reader. This means that only one item’s barcode can be scanned at a time, and a human must be present during the scanning process. This time-consuming, human involvement is not necessary for RFID.

Optical memory cards (OMCs) are another form of AIT. OMCs are the size of credit cards and use the same type of technology as CD–ROM products. Data are downloaded onto the cards in sequential order; once loaded onto the card, the data cannot be overwritten. In other words, portions of data cannot be erased (although the entire contents can be erased so that the card can be reused as if it were new). Additional data are loaded onto the card until its capacity is reached. These small cards can store over 2 megabytes of data. They are rugged, inexpensive to produce, and unaffected by environmental conditions such as moisture and heat. Smart cards (also called common access cards, or CACs) are similar to OMCs. Like OMCs, smart cards are the size of credit cards. While OMCs are used to capture information about supplies and equipment, smart cards are used to capture information about people. They currently have a data storage capacity between 16 and 32 kilobytes.

The contact memory button is another type of AIT, which is currently used by the Department of the Navy to store information about a major end item’s maintenance history. A memory button is a battery-free, read/write, identification device designed for use on components and equipment in harsh environments.

Understanding RFID

RFID is the most sophisticated type of AIT. To understand the complexities associated with RFID, it is best to introduce the rudiments of the science that makes it possible. Let’s start with a brief discussion of the electromagnetic spectrum.

According to the website of the Goddard Space Flight Center of the National Aeronautics and Space Administration (NASA)—

Electromagnetic radiation can be described in terms of a stream of photons, which are massless particles each traveling in a wave-like pattern and moving at the speed of light. Each photon contains a certain amount (or bundle) of energy, and all electromagnetic radiation consists of these photons. The only difference between the various types of electromagnetic radiation is the amount of energy found in the photons.

The spectrum of electromagnetic energy, from low energy to high energy, includes amplitude modulation (AM) radio waves, shortwave radio waves, very high frequency (VHF) radio waves (used by television), frequency modulation (FM) radio waves, ultra high frequency (UHF) radio waves, microwaves, infrared light, visible light (light that humans can see), ultraviolet light, x rays, and gamma rays. The electromagnetic spectrum can be expressed in three different ways: wavelength, frequency, and energy. AM radio (at the lower end of the spectrum) has long wavelengths (measured in meters—the distance between the crest of one radio wave and the next), low frequency (measured in cycles per second), and low energy (measured in electron volts). In comparison, gamma rays are at the highest end of the electromagnetic spectrum. They have short wavelengths, high frequency, and high energy.

Because so much technology is based on the electromagnetic spectrum, governments (including our own) have established guidelines to regulate portions of it. For instance, the Federal Communications Commission grants broadcast licenses to radio and television stations. Without some type of regulation, two radio stations in the same area, one a hard rock music station and the other a classical music station, might broadcast their different musical genres at the same exact frequency. Regardless of musical taste, the result would be unpleasant to hear.

RFID is based on the technology associated with the electromagnetic spectrum. Measured in hertz (1 hertz equals 1 cycle per second, where a cycle is the passing of one complete wave of energy), most RFID devices operate within a radio frequency of 124 kilohertz (124,000 cycles per second) to 2.45 gigahertz (2.45 billion cycles per second). RFID devices use the energy of radio waves as a basis for digitizing logistics information. The lower frequencies are less affected by metal and moisture than the higher frequencies, but the latter can be read at greater ranges. Radio wave readers (interrogators) emit radio waves to radio tags (transponders). The tags include both a mini-antennae and a computer chip; the latter contains digitized information about the items attached to the tags. The three types of RFID tags are active, passive, and semipassive.

An active RFID tag contains batteries. These batteries enable the tags to transmit information to a reader. A passive tag does not contain batteries or any other type of internal power source. It receives its energy from the reader, which emits its energy via radio frequency (RF) waves to the passive tag, which then uses the microchip’s antennae to convert this energy into electricity to transmit the information stored in its chip through its antennae back to the reader. A semipassive tag makes use of an internal power source that monitors environmental conditions and runs the chip’s microcircuitry. In order to conserve energy, many semipassive tags stay dormant until power is received from an interrogator. However, like passive tags they require RF energy transferred from the reader/interrogator in order to power a tag response.

Unlike barcodes, OMCs, or smart cards, RFID does not require human involvement in the scanning process. In fact, RFID tags do not have to be scanned via a line-of-sight process as do the other forms of AIT. Similar to audible sound (which is itself radio waves), RFID radio waves reverberate over a large area and can be captured by a reader even when the source of the radio wave is not within a line of sight. This means that the reader/interrogator does not have to be facing an item to sense it is nearby, as long as the item is located somewhere within the range of the interrogator. Moreover, the information transmitted by radio wave frequency can be captured quickly by the interrogator and the computer linked to it. The logistics information about thousands of different items can be captured within seconds, without human involvement.

Although several different types of active tags are used by DOD, the typical one, when compared to passive tags, is larger in physical size, costs more, stores more data, and transmits further. Active tags are about the size of a can of beer, cost about $70 each, contain 128 kilobytes of data, and can transmit information a radial distance (omnidirectional) of 300 feet if unobstructed. Passive tags are much smaller (about the size of a postage stamp), cost as little as $0.25, store as little as a few bits of data, and must be read within 3 to 10 feet.

Active RFID Tags

The most sophisticated active tags can transmit data directly to satellites, which then relay information to appropriate wide-area networks. Because of the high cost of these tags and the expense associated with using satellite telecommunications, these types of tags are reserved for tracking time-sensitive items, such as critical ammunition or perishable items. However, these tags are being used more often, and associated costs are expected to drop.

In Iraq and Kuwait, logisticians are using satellites to track the locations of specially equipped containers that have been outfitted with “AXTracker” global positioning system (GPS) tracking devices, which are manufactured by a high-tech corporation called Axonn. These self-contained devices, which are 9 inches by 6.25 inches by 1 inch, can be attached to containers easily. They send signals to low Earth-orbiting satellites, which relay a container’s current GPS location and other information such as ambient temperature, humidity, and whether or not the container is stationary or moving. The satellites then transmit the information to ground-based computer gateways that pass the information to web portals. The AXTracker has specially designed batteries that can last from 3 to 18 years, depending on how often information is relayed to the satellites.

Active tags—the type typically placed on railcars, major end items, 20-foot containers, and 463L pallets—have their own unique tag identification numbers. When active tags are placed on 20-foot containers, they normally do not contain very much data on the items within the container. However, AIT systems and wide-area networks are being designed to correlate an active RFID tag number with the physical location of the tag, the time the tag was read, and the contents of the container. This is currently accomplished at DLA container and consolidation points and DOD-run seaports in the United States.

Active RFID readers normally are positioned at transshipment points, at preselected checkpoints along a designated route, and at end unit locations. Unlike passive tags, some of the data stored on an active tag can be changed using radio wave frequency transmission. Usually, however, a person loads the appropriate information onto a tag at a computer docking station. The loading of digital information onto a tag is called “writing” or “burning.”

As with most technology, active RFID tags are not without certain problems. Their batteries go dead; the more frequently data are exchanged between the tag and the reader, the sooner this happens. For example, if an active tag on a 20-foot container being hauled by a tractor passes by a fixed interrogator at 20 miles per hour, the tag will be read only once. However, if the tractor-trailer happens to be parked for several days near that same interrogator, the tag will be read numerous times and its battery life will be significantly reduced. Furthermore, because of the omnidirectional nature of the RF transmissions, the same active tag may be read by two separate interrogators at the same time.

Passive RFID Tags
Unlike active tags, passive tags are being designed to be an integral part of nearly every piece of equipment
and of all but the smallest or most inexpensive supplies. As technology improves, the cost per tag is expected to drop below $0.05. Active tags are normally placed on items at transshipment points. However, passive tags are assimilated into items at the point of manufacture. Passive RFID tags in transit or in storage can be read by readers at a rate of more than 1,000 tags per second.

Passive tags do not contain very much logistics information; they are similar to that of a license plate in that they contain only a few alphanumeric characters. A passive tag, like a license plate, is used as the key for obtaining additional information. For example, a police officer who stops a car for speeding can simply call police headquarters and report the alphanumeric characters of the license plate. Headquarters then would use the license plate number to access its database and uncover all types of information pertaining to the car and owner, such as vehicle registration data and arrest records.

RFID Shortcomings

RFID technology, particularly passive technology, is still relatively new and developing. Many of the active tags and the interrogators designed to read them are not interoperable with the passive tags and interrogators. It is almost as if there are two separate systems. Moreover, a comprehensive architecture for the AIT systems and the wide-area networks associated with RFID has not been designed on a global scale yet, although a great deal of progress has been made in areas where U.S. forces have an established presence.

A major weakness of RFID is that the technology can be interrupted by the enemy. Interrogators placed along delivery routes are easily seen and sabotaged, and their very emplacement indicates the location of a main supply route. Furthermore, it is unlikely that rapidly advancing forces will have the time and wherewithal to establish interrogators along the route of troop advance. Data-read capability is also an issue. The placement and positioning of the tags and the interrogators makes a difference as to whether or not the data are successfully transmitted.

Active tags are more susceptible to enemy interference than passive tags for several reasons. First, since active tags transmit data over longer distances, the enemy has a better opportunity to corrupt or impede the transmissions. Second, active tags contain more information for the enemy to corrupt than do passive tags. Lastly, the enemy could actually change the data on active tags that have read/write capability since most tags currently in use are not encrypted.

In some DOD experiments, passive tags placed on the inside of multicontent pallets were not picked up by the interrogator. This could have occurred because certain radio frequencies do not penetrate easily through liquids or humid conditions. For instance, paper products have high moisture content, so their passive tags do not always enable the capture of complete and accurate information. Similarly, the RF waves associated with passive tags do not readily pass through dense objects or metals. Data-read also depends on the quantity of tags being read by a single reader, the speed of the tags passing by a reader, and the distance between the tag and the reader. The more tags being read, the greater the speed of the tags in motion, and the greater the distance the tag is from the reader, the less reliable the reads will be. Probably because passive RFID technology is in its nascent stage, some studies indicate that the low-cost tags can be damaged during production, which frequently happens when microchips are attached to the mini-antennas. Tags can also be damaged when logistics data are written to them.

Additional issues are associated with RFID technology. Tags that use a frequency of 433 megahertz can interfere with military radar. The electromagnetic energy of RF can adversely affect people, ordnance, and fuel. RF transmissions at 2.45 gigahertz excite water molecules; not coincidentally, this frequency is used by microwave ovens to heat food.

Like the United States, foreign governments regulate frequencies within their airspace. This means that DOD must obtain permission from foreign governments for deployed U.S. forces to use certain portions of the electromagnetic spectrum. Currently, no international agreements stipulate electromagnetic frequencies for RFID. The spectrum being used for RFID ranges from 860 to 960 UHF within the United States, Europe, and Japan. However, there is a growing global acceptance of designating a few frequencies exclusively for RFID use.

Regardless of the method in which it is captured, logistics information pertaining to asset visibility must be accurate and complete. Commanders and logisticians must firmly believe that the information they retrieve about items in storage or in transit is reliable. After all, supply and equipment readiness is an integral part of combat power.

Lieutenant Colonel James C. Bates, USA (Ret.), works for Alion Science and Technology and serves as a sustainment planner for the U.S. Joint Forces Command, Standing Joint Force Headquarters (Standards and Training), at Norfolk Naval Base, Virginia. He is a Certified Professional Logistician and a graduate of the Army Command and General Staff College and holds an M.B.A. degree from the University of Hawaii. He can be contacted at James.Bates@jfcom.mil.