Tune to a new frequency
Radio-frequency identification could improve supply chain efficiency, but it may take a while
By Tony Baer, contributing editor -- Manufacturing Business Technology, 11/1/2003 7:00:00 AM
"Bar-code technology is like employing an automated typist; it simply lets you record bits of information and store it in an application's database for possible later use," explains Noah Dixon, RFID strategist for RedPrairie, a supplier of distribution-management software that is working with several customers on RFID pilots. "RFID, on the other hand, provides streams of data that can be differentiated and interpreted before being passed off to an application."
RFID—which uses radio waves to pull data relating to an object, or an individual, from specially designed ID tags—represents a definite advance beyond today's ubiquitous bar codes. Some industry experts believe it's also the missing element in what has been a decades-long search for a way of actually squeezing excess inventory out of supply chains.
The first advantage of RFID over bar coding is that RFID does not require direct contact or line-of-sight scanning. It uses an antenna and a transceiver—often combined in a single unit—to read the RFID tag, which in technical terms is called a transponder.
The antenna uses radio-frequency waves to transmit a signal that activates the transponder. When activated, the tag transmits data back to the antenna. The data goes to a programmable logic controller, telling it that an action should occur. The action could be as simple as raising an access gate—which is what happens when cars equipped with RFID tags pass through a tollbooth—or as complicated as interfacing with a database to carry out a monetary transaction.
Exciting possibilities
The latter capability is what excites manufacturers, distributors, and supply chain management experts about RFID. But the technology's seemingly boundless potential also is hindering it from gaining widespread commercial use, primarily because many of its proposed uses are considered either too impractical, too impersonal, or simply too difficult to execute.
For instance, Unilever, the global consumer goods giant based in The Netherlands, dreamed up the concept of a home inventory management system that can be managed via a PDA as part of a "smart kitchen" pilot it developed while exploring possible uses for RFID technology. The long-term vision for this project is the creation of a global information network that triggers the movement of specific goods from the manufacturing plant to the retailer, and finally to consumers' kitchens, all on the basis of instructions relayed by RFID tags. But Rick Raber, RFID project manager for Northern Apex, an Indiana-based engineering firm that helped Unilever create the demo, says the common initial reaction is that the idea is "way too Jetsons."
Apparel manufacturers have toyed with the idea of putting RFID tags inside individual garments so they can track exactly which items are selling fastest in various parts of the world, but consumer focus groups tend to view that as too Big Brother, even though manufacturers say they can protect consumers' privacy by adding a means of de-activating the tag once a garment leaves the store.
These objections are partly why few, if any, commercial RFID projects involve what is known as item-level tagging. In most business applications, tags are attached to cases, pallets, or trucks carrying goods, which allows tracking the movement of entire orders, but not necessarily the individual items within an order.
While item-level tracking could enhance manufacturers' ability to track actual consumer demand for products, there is some question about its broader value as a supply chain management tool. "Today, if I scan a bar code, I know it's a case of Wisk, but I don't know when it was made," observes Simon Ellis, supply chain futurist for Unilever. "RFID gives me the chance to know which case of Wisk it is. The question is whether I really need to know all that."
"If you don't watch out, you could get overwhelmed with the information you can acquire via RFID tags," adds Mike O'Shea, Auto-ID and RFID strategy director for Kimberly-Clark, Dallas, another major consumer goods manufacturer that has been experimenting with the technology.
Store of tomorrow
A glimpse of RFID's potential is on display in Rheinberg, Germany, where enterprise software supplier SAP is working with Metro, one of Europe's largest retailers, and several major consumer goods manufacturers to outfit a supermarket with RFID technology. Scanners are placed in the packing, shipping, and receiving areas of a Metro distribution center to track which pallets are delivered on which trucks. At the store, RFID tags on shelves send signals when it's time to replenish certain items. Shopping carts also have tags, and, in a few cases, item-level tags are placed on specific packages of product.
Alexander Renz, SAP's director of Auto-ID, says the goal is to test the usefulness of RFID for improving supply chain accuracy, while also discovering new services that could stimulate business, such as helping consumers locate items in the store, cross-selling related items—even eliminating checkout lines by scanning and totaling customers' purchases as their shopping carts whiz by RFID readers.
"The Metro store of the future has combined all these technologies into a real-world store where people go shopping every day," Renz says. Still, there are numerous obstacles—many of them technical in nature—that must be conquered before every neighborhood grocery store is equipped in this fashion.
"While the Holy Grail of RFID is that we will magically scan data related to large numbers of items as they pass through doorways, the laws of physics preclude that," says David Hardy, director of research and development for Acsis, a supplier of data collection hardware.
For starters, the "noise" generated by electric motors and electronic controllers in the average factory or distribution facility could interfere with radio frequencies, while the placement of metal posts could create radio dead spots. Certain products can pose challenges as well. For instance, while paper and plastics should be easy to scan, metal cans might reflect RF waves, and liquids absorb them.
Managing all of the data generated on an RFID network could be an even bigger challenge. This issue comes into play on two levels: when deciding how much detail to collect on specific supply chain events, and when evaluating if and when item-level tagging makes sense.
All news all the time
Deciding what event details to collect is important because RFID readers, unlike bar codes scanners, are radios that stay on all the time. "You need a true message-based architecture that listens for data, buffers it, and acts as a traffic engineer, with the ability to decipher all those messages and know how to dispose of them," explains Paul Weiss, chief technical officer for V3 Systems, a logistics-management software provider.
Adam Bartkowski, CEO of Apriso, which sells software for managing both shop-floor and supply chain processes, suggests that the event data an RFID system collects should correspond with specific business rules. "It should adhere to a rules-based structure that addresses the question of how frequently you really need to monitor events."
Raymond Blanchard, who directs business development for SAP's manufacturing solutions group, notes that the Metro store of the future will employ multiple brands of RFID readers that have different methods of filtering data. Some will only accept data every 4 to 7 milliseconds. Nevertheless, it would present a whole different problem if the Metro project progressed to the point that it was necessary to aggregate and analyze data collected by the different readers. "That's an integration nightmare that standards must solve," Blanchard says.
None of these problems are insurmountable, particularly in the minds of RFID proponents. Companies like Linx Data Terminals, a vendor of RF systems for collecting labor and other information on shop floors, have an inside track on developing terminals that can deflect possible interference in a distribution center.
Vendors also are designing distributed RF networks in which data is collected in a small, semi-closed environment before being passed off to a middleware program that transforms it into the proper format for merging with data from other closed-loop systems.
"Software architecture would be based on edge servers and middleware, which cleanses data before it gets to the database," explains Chris Clauss, worldwide auto-ID leader at IBM.
The cost factor
Despite concerns about privacy and data overload, tag costs seem to be the biggest factor preventing companies from being more aggressive at launching projects involving item-level tracking. Gartner, the Stamford, Conn.-based IT research and consulting firm, says major manufacturers such as Procter & Gamble have indicated that they won't consider item tags until costs drop below a nickel per unit. By comparison, Gartner estimates that case or pallet tagging would become cost-effective at 10-cent levels.
Wal-Mart, Bentonville, Ark., apparently believes case and pallet tagging is affordable now. The retail giant has told its top suppliers that it wants RFID tags on all cases and pallets delivered to its distribution centers by 2005. Some industry observers believe Wal-Mart's edict will speed the development of RFID solutions, just as a similar mandate issued by this retail giant a decade ago inspired the widespread use of bar codes.
"You'll see dual systems," Acsis' Hardy says. Not surprisingly, companies that provide equipment for printing labels, such as Zebra Technologies, are offering "smart labels" that embed radio tags inside standard bar-code labels. Matt Ream, senior product manager for Zebra, says these labels also can be a source of security. "You'll always need a human-readable label just in case RF systems fail," he says.
Despite the obvious pitfalls, early adopters are bullish on RFID technology. "Wal-Mart's announcements are not just a mandate; they are an endorsement of the technology," insists Kimberly-Clark's O'Shea.
Ellis, the Unilever RFID futurist, points out that bar-code technology faced similar skepticism in its early days. "It took 20 years to get 90-percent penetration rates for bar codes," he says. "RFID shouldn't take that long."
