The following section describes RFID technology at an overview level. It is intended to give readers who are unfamiliar with RFID or just need a brief refresher, a thumbnail sketch of the technology's basic functionality and its intended markets, as well as, an understanding of its relevant implementation and political issues. For more detailed information on such topics as RFID architecture and its associated standards, please investigate the resource links on this page.
The basic concept of Radio Frequency Identification (RFID) is that a query is sent out over radio waves ("What or who are you?") and then a subsequent reply is received ("This is what or who I am").
When you hear the term RFID used today it most commonly refers to the burgeoning
business application for managing and tracking supply chains, especially in the
materials, manufacturing, and retail industries. The supply chain business objective is to
use RFID technology to not only take the just-in-time inventory concept to its next
performance level but to support additional information functions ... from streamlining the
product recall process and reducing theft and fraud, to further improving the type and
quantity of product marketing information.
RFID technology was invented in 1948, but it was not mainstreamed for commercial applications until the 1980s. One of its first known applications was during World War II, when it was used by the British radar system to differeniate between German aircraft and their own aircraft with attached radio transponders.
Only recently, due to technology advances, have the price points dropped to where RFID is now feasible for companies to adopt. Wal-Mart was among the first commercial enterprises to select RFID technology to achieve improvements in the inventory supply process and theft control. Wal-Mart started the process of implementing RFID throughout its retail distribution chain by requiring its top 100 suppliers to use RFID tags by year-end 2004 on the pallets and cases they shipped to Wal-Mart.
However, in 2004 due to the status of standards and limited tag production capabilities, manufacturers of RFID tags were unable to meet the volume of demand within Wal-Mart's desired time frame. As the standards further evolved, production capabilites expanded and price points kept dropping, Wal-Mart's implementation is now in full swing with most of their top suppliers delivering RFID tagged pallets and crates to all of Wal-Mart's distributions centers.
RFID use has started to penetrate other corporate supply chains, but in many cases these implementions are closed loop systems. In other words, the RIFD is deployed within a particular corporation's supply chain, but not between them and their business trading partners. There are other applications which are also limited to deployment withing a single organization, such as tracking tools so they aren't left behind in the airplane after maintenance.
The underlying technology architecture of RFID is based on these components:
tag and its associated data structure
reader with antenna and the reader's associated software
communications protocol suite
database - data synchronization
A reader can be either stationary in a fixed state (e.g., mounted above a conveyor belt) or mobile as in a handheld device or attached to a forklift. The tag is a miniature chip with an affixed radio antenna. The tag is attached to an item or its packaging. A radio wave signal is transmitted between a reader and tag to communicate an Electronic Product Code (EPC). The EPC is used to uniquely identify the pallet, case, or item.
An EPC can identify anything in a supply chain hierarchy. For example, a single EPC could designate a case of Purina® DogChow® purchased by PetSmart, or it could designate the specific 25-pound bag of Purina® DogChow® Healthy Morsels purchased 09-25-06 at PetSmart by Joe Smith at 2555 Main Street in Chicago, Illinois, tel. 312-222-2222. At the more detailed individual item levels, privacy issues arise that are not explored within this high-level technical description.
In the future, it is envisioned that there will be even more information tracked via the EPC by using the EPC Information Services (EPCIS) "discovery services" into the Object Name Service (ONS). The ONS will redirect a query on a specific EPC by resolving the EPC code into an Internet address, where considerable detail may be stored by whomever has had the item in its "custody". What details might be tracked and which "custodians" actually elect to (or are required to) write those details are undetermined at this time. Verisign won the contract that was awarded by EPCglobal to manage the ONS root. (It is of interest that Verisign also manages the Internet's root DNS.)
Currently EPCIS is in its infancy and "discovery Services" are just a glimmer in the eye of the RFID developers. EPCIS is a standard interface definition for accessing EPC related data such as event data (e.g.tag read at timestamp 17:31:002 at loaction XYZ), serial-level attributes (e.g. date of manufacture, data of expiration, adn etc.) and location Master Data for site (i.e. identifiaction of reader locations within a store or distribution center).
There are currently two types of tags: passive and active. Passive tags have no directly associated power source, while active tags do. Passive and active tags can be either class 0 (read only) or class 1 (read/write) tags. The approved radio frequency range for RFID applications is 900MHz for Class 0, and either 13.56 MHz ISM Band or 860-930 MHz for Class 1, depending on the strength of signal required.
In the case of a passive tag the reader initiates communication via a radio signal strong enough to enable the tag to "answer" the reader with a return radio signal carrying information regarding the item to which it is attached. In the case of an active tag either the tag or the reader can initiate communication. Further, active tags allow for a greater distance between the reader and the tag. It should be noted that with the recent improvements in passive tag technology the distance limitation is decreasing - in other words some passive tags are croosing over into read distance ranges previously only supported by active tags (for more information check out Intel's R1000 chip.
The transmitted radio frequencies are used to exchange a very brief and small amount of data,
the EPC, which is then used to map to information regarding an object’s unique
characteristics such as:
• This is where I originated: "manufacturing source"
• This is what I am: “a sepcific pallet carrying boxes of paper towels” -or-
• This is what I am: “a specific box of paper towels”
The use of the word"specific" in the above example refers to the fact that we are identifying a particular instance of an object - hence a unique identifier (EPC) is used for each instance. This is also referred to as following a serialized concept similiar to a serial number you have on your regfrigerator or desktop computer. This technique of using unique EPCs is consistent with the original developers concept. However, it should be noed that there are cases were dulplicate EPCs are being used, particulary at the item level when used as a direct correlation/mapping to an existing bar code on a item.
• This is my route tracking: “my mode of transportation; my origination point, my
arrival and departure time and location of every stop along the way, and the time
I arrived at this final destination point”
• This is my purchase and restocking information: “I was purchased by so and so, on this date, at this price, and I have now left the building – please restock this item”
At this time, tagging is not occurring at the item level in most cases but rather at the pallet and case level. Tagging at item level is being done in some high-price ticket items. Best Buy, for example, is beginning to use item-level tagging for items like computers. The plan is to migrate down to the item level in the entire supply chain as it becomes economically feasible.
The data structure for the EPC has been defined by the EPC global organization. The data structure lays out the data fields and what information they carry, the size of the data fields and how they are digitally encoded.
A reader has a field (a distance range) within which it can query via radio waves for whatever tags may be present. The reader follows a protocol that is intended to enable it to avoid duplicate reads but capture all tags present within its range.
There is a large variation in reader capabilities, ranging from how many tags a reader can capture within a specific time period to more complicated tasks like filtering and communicating with a product database. Readers have to be matched to tag type: active or passive class: 0 (read only) or class 1 (read/write), gen 1 or gen 2 tags. Some readers can capture multiple tag types. Readers communicate with tags over radio waves and by following a specific RFID communications protocol.
The implementation challenge with readers is determining how many you need and their ideal placement in your physical facility. Poorly placed readers can result in creating duplicate reads of tags when their read range overlaps (some readers have special edge software to address this problem). Or, you can end up with the opposite result with missed item reads when they are located too far away from the tagged items.
Current: The EPC and ISO have standardized the first two layers of the communication protocol stack between the readers and the tags. These two layers include the local wireless communication that occurs between a reader and the tags within its read field. The first layer standard is the physical, which describes the specific radio frequencies and whether tags and readers are communicating in half or full duplex mode. The second layer, referred to as the data link layer, has been standardized based on a slotted Aloha scheme.
Middleware standards have been defined to support temporary collection of event data for filtering and consolidating the EPC data comming from the readers. This standard is called Application Level Events or ALE. Another communication standard has been defined for the readers in terms of how they capture and commuicate event data from tags and sensors - called the Reader Protocol (RP).
Communications between readers and in-house databases are up to individual implementations. This is also true for inter-company communications.
Current: Today there is no such thing as an "RFID network" except in the local/private sense, the extremely local communications between tags and readers, and subsequently between the reader and the in-house database system. Although it hsould be noted that the ALE middleware standard operates at that first edge between the reader and the in-house database. RFID network links include the local wireless (RF) network between tags and readers, between readers and the internal company database systems (e.g., ERP or WMS in the retail industry) and in the inter-company network links which today are individually negotiated between business partners.
Future: The planned-design intent behind RFID architecture is to ultimately standardize communications between
entities based on an Internet architecture-based construct. There is considerable work that needs to be done
to make this vision a reality.
Current: Today, retail companies and their suppliers are using RFID very much like they have been using UPCs. Suppliers/manufacturers provide the retail companies with a listing of EPCs and what pallets/cases of product they represent. The retail companies typically then feed that information into their existing in-house database systems (e.g., WMS, ERP). In other words, today there aren't any real world databases built on the RFID Internet based architecture proposed by EPCglobal.
A very real and major challenge facing the manufactures/suppliers and the retail distribution companies is keeping data synchronized. In fact, maintaining data synchronization within an indivdual company is challenging enough --- between companies is orders of magnitude more difficult.
The UCC (Uniform Code Council) is working to address this problem in today's world of UPC bar codes via the UCCnet. The following is a quote from the UCCnet Web page: "UCCnet is a subsidiary of the Uniform Code Council, Inc.® (UCC® ), a not-for-profit member organization of GS1 (formerly EAN International) that is dedicated to the development and implementation of standards-based, global supply chain solutions. UCCnet is a GDSN-certified, U.S.-based data pool that offers data synchronization services that enable trading partners to exchange accurate, standards-compliant data."
It is hoped that some of the lessons being learned in that effort will be able to flow through to the RFID data services architecture being defined by EPCglobal.
Future: An RFID tag will be able to generate and even carry on its chip event/sensor data that could be used in a variety of ways.For instance, sensor data could include temperature ranges experienced enroute by perishable food items like eggs. In fact, RFID tags will enable/create volumes of data throughout the supply chain.
One of the daunting challenges will be how to make that data useful to businesses and consumers in the supply chain.
Air Interface and Data Structure Standards for Tags and Readers
The two dominant standards bodies for definition of air interface and tag data structure standards for RFID are EPCglobal and the International Organization for Standardization (ISO) [www.iso.org]. Typically the “industry motivated” group (in this case, EPCglobal) gets their standards drafted first and then has the job of persuading ISO to adopt the standards they propose.
Specifically, EPCglobal has laid out the following tag interface specifications:
EPC tag Version 1.1
900MHz Class 0 Radio Frequency (RF) Identification
13.56MHz ISM Band Class 1 Radio Frequency (RF) Identification
860 MHz -930 Mhz Class 1 Radio Frequency (RF) Identification
ISO standards roughly correspond to the standards listed above. In December 2004, EPCglobal ratified its UHF generation 2 standard (for passive tags class 1). The ISO “version” of this is the draft standard ISO18000-6. One field is still somewhat in dispute between the two standards bodies. There are several improvements and advantages to the new gen2 passive tag specification.
The new spec improves performance relative to data rates and additional security, including a kill feature (important for privacy) and closer alignment with ISO standard 18000/6 part C. In addition, it will be able to operate globally, not just domestically (like current passive class 0,1 tags). In addition to EPCglobal and ISO, given the Internet component of the architecture relative to the ONS one must also look at published RFCs by the IETF (the Internet standards body).
The standard tag identifier in the supply and distribution chain until recently has been the ubiquitous Universal Product Code (UPC) bar codes – the black and white stripes that cashiers scan to read a price. The difference between them and an Electronic Product Code (EPC) is that UPC codes require line of sight between the bar codes and the scanners. UPC technology is based on a system of bar codes that, when scanned by a laser, identify the item being scanned (e.g., a loaf of bread, a music CD, a rifle, a specific book title). UCC (Uniform Code Council) is the organization in charge of assigning UPC manufacturer codes. (This can be construed as similar to how Bellcore was responsible for assigning phone number groups to RBOCs and LECs.)
RFID tags, on the other hand, are not restricted by line of sight. EPC codes use RFID tags attached to an item, which use radio frequencies to communicate to an RFID reader. What this means to a business is that a product can be tracked from the time it is tagged (e.g., at the factory) until the time it is ultimately discarded (e.g., at a landfill). Because the technology does not require line of sight in order to identify an object, materials handling is greatly reduced.
Similarly to the UPC, the government (Department of Defense, DoD) has been using Unique Identifier Codes (UID) to identify those same types of items plus others such as, guns, ammunition, tanks, etc. Companies like Honeywell, for example, supply product to both the commercial marketplace as well as to the military marketplace. Hence, previously Honeywell would have one “manufacturer’s code” with DoD (a UID code) and a different one from UCC (a UPC code). Now that the DoD has also embraced the new EPC standard, Honeywell and other coprorations will be able to use the same EPC for both the government and commercial retailers.
Current: Simply stated RFID can be an effective replacement technology of UPC bar codes. Because no line of sight is required. RFID can be used more easily by retail chains than today's bar code.
Future: RFID has an almost unlimited potential in what it could track. It could track an item throughout its lifecycle. Such applications might include keeping records of maintenance and repairs or tracking personal data such as medical records and passport information. A tag could be attached to an inanimate object or embedded under a person's skin.
But the potential of RFID, depends on how well the standards definitions are adopted, their realization into actual equipment components, how expensive they are, and how they are received by the purchasing public.