The essence of RFID (radio frequency identification) technology may be invisible, but that won't stop all eyes in the supermarket industry from turning toward Dallas next month.
That is where Wal-Mart will fire the first shots in the RFID "revolution," which has made RFID one of the most talked-about technologies in many years. The retailing giant will start testing the technology to track inventory at a Dallas-area distribution center, with pallets and crates of goods from several suppliers "tagged" to identify themselves wirelessly via RFID rather than with handheld scanners.
The revolution is actually a long time in coming. RFID itself has been in commercial use since the 1980s, including Mobil's pay-at-the-pump Speedpass device and automatic toll payment systems like EZ-Pass in the New York area. Its current incarnation, based on the EPC (electronic product code), has been in development for five years at MIT, where an army of academic and industrial researchers pioneered the low-cost microchips and readers that will be the RFID weapons of choice for retailers and manufacturers. The EPC is a digital numbering system whose unique identification information is carried in the microchip part of the tag and read by a nearby reader.
Since last year, Wal-Mart has been marshaling the troops with its decree that by January 2005, its top 100 suppliers must deliver pallets and cases with RFID tags; the rest of its suppliers would follow suit by December 2006. (For the latest RFID news on Wal-Mart Stores, see Page 1.) Albertsons and Target joined the fight recently with announcements that they, too, will start adopting the technology next year. Metro Group, Germany's largest retailer, made a similar announcement at the National Retail Federation show in New York in January.
Others are now mobilizing. "I would not be surprised to see other major grocers come out with mandates this year," said Christopher Boone, a retail industry analyst at IDC. A recent report by A.T. Kearney and Kurt Salmon Associates, "Connect the Dots," concluded that "pallet and case level applications of EPC will be widespread within three years." (The report is available at www.fmi.org. Food Marketing Institute is one of its three sponsors along with Grocery Manufacturers of America and National Association of Chain Drug Stores.)
Other think tanks are chiming in. Forrester Research estimates that a typical supplier will spend $9 million in 2004 to comply with Wal-Mart's mandate alone. Venture Development Corp. calculates that worldwide spending on RFID will reach $1.65 billion next year and $2.7 billion by 2007. Not surprisingly, vendors selling hardware, software or consulting services are rushing to offer RFID-compatible products. Industry conferences to discuss RFID routinely draw crowds from across the country and the world (see stories, Page 1 and at right, respectively). Informally, many supermarket executives said it is one of the most intriguing technologies they've encountered.
When pressed for details, however, most grocers and suppliers remain mum. Albertsons, Shaw's, 7-Eleven, Ahold USA, Associated Grocers and Tesco all declined through spokesmen to discuss their dabblings with RFID. Some fear that publicity invites backlash from RFID critics worried about privacy abuses. Others stay silent for competitive reasons.
Of course, it's still early in the game, and most companies just don't know what they're going to do with the technology. "There are very few people who know what this is going to look like," said Christine Overby, a consumer goods analyst with Forrester Research.
"A lot of people are taking a wait-and-see approach," said Michael Sansolo, senior vice president, Food Marketing Institute. Most grocery executives are enthusiastic about RFID's potential to combat theft, track deliveries, monitor food freshness and more, Sansolo said. However, they are acutely aware that RFID is still in its infancy. "It's all a lot of 'ifs' at this point," he said. "It hasn't been demonstrated yet."
Thinking Big Thoughts
The largest challenge for any supermarket or supplier right now is to determine what business processes it wants RFID to track. The rote answers are usually lofty visions about curbing theft or reducing overstock, but that means little to supermarket chief information officers trying to select vendors or develop pilot tests. Faced with a very blank canvas, most want to see what picture Wal-Mart and Albertsons paint before attempting anything themselves.
Wal-Mart plans to use RFID to monitor shipments of goods as they arrive from suppliers and move around its distribution centers, with the ultimate goal of curbing out-of-stock shortages at local Wal-Marts and Sam's Clubs. Assuming its tests in Dallas go well, the company will roll out RFID to other distribution centers in Texas and the rest of the country next year, although spokesmen will not disclose any specific timetable.
Albertsons, far newer to the game, is more reluctant to talk about its plans. The company said it is already testing RFID on cases and pallets of grocery goods, but won't identify how large the tests are or which suppliers are participating. The 2,300-store chain wants its top 100 suppliers to start delivering RFID-compliant pallets and cases of goods by April 2005.
Other supermarkets still hesitate to commit precious IT budgets to the technology. An emerging rule of thumb: The smaller the company is, the more uncertainty it has about how RFID can improve operations. "They're asking themselves, 'What am I going to do with these tags? Why are they better than bar codes?"' said Jeff Woods, an RFID analyst with the Gartner Group.
For many grocers, the answer might be that tags are not better than bar codes. According to various estimates, the tags cost around 50 cents each on the low end. While that is practical for pallets and cases of goods, it is far too expensive to tag individual items (and the technology still works poorly around metals, liquids and meats anyway), which is where most small grocers stand to reap any business benefit.
"The issue becomes what value it has in the store," said Jerry Morris, vice president of TCI Solutions, which makes software for the grocery industry. He agreed that RFID is "wonderful" for shipping logistics, but reported hardly any interest from clients that do not have their own distribution operations.
Small grocers "are dealing with bars of soap and cans of soup," he said. "That's a very cost-conscious decision."
The Aholds, Albertsons, Wal-Marts and other large chains of the world, on the other hand, can stand to reap significant new efficiencies from RFID, assuming they can map out what data to track.
"It totally depends upon what your company is trying to achieve," Wal-Mart spokesman Gus Whitcomb said. "It's entirely possible to over-complicate things and store unneeded data. Therefore, it's very important to understand your business and make good decisions about what data is important and what data isn't."
Accuracy of data is also important. RFID could just facilitate more errors if base data is not cleaned up and synchronized with trading partners as part of the Global Data Synchronization (GDS) initiative. "Companies must pay the price of GDS implementation before the collaborative benefits of EPC can be realized," said the "Connect the Dots" report. "Without GDS, EPC technology represents nothing more than an expensive bar code."
In addition, when RFID does get to the individual item level, retailers will need to address latent consumer concerns about privacy (see story, Page 60).
Question of Standards
Gregg Maggioli, president of Indianapolis-based MagTech Systems, an IT consulting firm, has a handful of clients exploring RFID systems. He said many worry that today's lack of industry standards could leave them stranded if they choose poorly.
"If I'm a CIO, I'm hesitant to invest in anything until it's standardized," he said.
Maggioli installed one RFID system last year for Paramount Farms, a California supplier of pistachio nuts and almonds. The company used readers and tags from Intermec Corp. and custom-developed programs based on Microsoft software to retool its system of tracking harvest loads. That cut order processing time by 60% and leased-trailer usage by 30%, enough efficiency that Paramount scrapped an $80,000 expansion of its scale house.
Paramount, however, uses a "closed" system -- one vendor providing all the equipment for functions that happen solely on Paramount premises. The Wal-Mart mandate pushes an "open" system, where everyone selects his own RFID vendor and the equipment operates in harmony. Such orchestration looks good on paper, but the actual symphony might be out of tune when it starts next year.
"This will be a great year for closed systems, and a pretty good year for open-system prototypes," Maggioli predicted.
EPCglobal, a subsidiary of the Uniform Code Council and EAN International, is racing to deliver industry standards for those open RFID systems, based on the EPC. The group ratified the first generation of RFID tag specifications last fall (known as Version 1.0), which serves as the platform for pilot tests by Wal-Mart, Gillette, Albertsons and others experimenting with RFID.
EPCglobal is already working on a second set of specifications, scheduled to be ratified by this fall, to improve the performance of RFID tags. The next generation of tags should have faster data-transfer speeds, longer range for RFID readers, and compliance with various power-supply standards around the world.
A critical part of drafting those new standards, EPCglobal officials said, is feedback from early-adopters.
"We're taking our time and being very careful," said Sue Hutchinson, a technical expert with EPCglobal. "Our user community is now in a better position to describe a good working protocol."
Erik Michielsen, senior analyst at ABI Research in New York, opined that the second generation of standards will sooth the nerves of skittish CIOs. He anticipated that many more businesses will tackle RFID more directly by the latter half of this year.
Gartner Group's Woods, however, fretted that EPCglobal will be the battleground for equipment vendors jockeying to make their platforms the standard for everyone. "The standards are very much in flux now," he said. Woods asserted that many CPG companies worry that they will embrace one technical standard for Wal-Mart right away, only to see another one supplant it in the future.
Manufacturer Movement
Supermarket CIOs have at least one advantage: They can stall for time until standards are settled. Large suppliers are in the more difficult position of having to embrace RFID immediately to satisfy Wal-Mart and a growing list of others. They must either adopt untested software right away, or wait until the bugs are solved by mid-year and race to install a system in the second half of 2004. Neither choice is comforting.
Grocers worry that the Wal-Mart mandate has made suppliers too busy with RFID to concentrate on anything else. An executive at one regional wholesaler described it as "everyone running around trying to please Wal-Mart."
Proctor & Gamble, which has helped underwrite RFID research since 1999, will start its first tests of the technology later this year. "We are certainly committed to doing a pilot with Wal-Mart, and there are several other pilots on the table," a company spokeswoman said (see story, Page 60).
Gillette has run a pilot at a distribution center in Fort Devens, Mass., since last year to test the performance of RFID tags it purchased from Alien Technology. Spokesman Paul Fox said results are "highly encouraging," and the company has teamed up with software vendors including Manguistics and Sun Microsystems to gauge how Gillette's IT systems must be overhauled to manage RFID-generated data. That, Fox said, "is still very much a work in progress."
Wal-Mart will develop its own applications to manage RFID data, true to its policy of developing all IT systems in-house. Smaller grocers will face a dizzying choice of vendors and consultants, from giants like IBM and Microsoft down to tiny shops like Maggioli's, claiming expertise in developing RFID systems. EPCglobal has promised a certification process to validate those claims, which will start with hardware vendors and eventually expand to software vendors and consultants as well.
Another fear small grocers have is that the cost of RFID will be passed onto them, since Wal-Mart has made it clear that it will not accept price-hikes from manufacturers. Also, RFID is going to be expensive. At 35 cents a tag, a manufacturer shipping 10 million crates will pay $3.5 million just for tags, let alone affixing them or managing new data. Industry estimates are that Wal-Mart receives upwards of 1 billion cases and pallets of goods annually. Analyst Michielsen said much of that cost will eventually be recouped through better efficiency, but the first few years are likely to be turbulent.
Ultimately, however, few people doubt that RFID will be a formidable new tool to manage inventory and smooth bumps in the supply chain. The challenge in 2004 will simply be to devise experiments and start conducting them to see what RFID can do.
Woods from the Gartner Group estimates that 50% to 70% of all first-wave RFID implementations will be duds. "People have to be willing to fail here," he said.
