Simon Ellis can see the future. Indeed, that's a large part of his job description. As the inventory and logistics guru at Dutch consumer-products giant Unilever, Ellis carries the peculiar title of Supply-chain Futurist. Thus, he spends a whole lot of time thinking about things like ERP (enterprise resource planning), WMS (warehouse management systems), DDSN (demand-driven supply networks), and OMTA (other maddening technology abbreviations).
These days, Ellis especially ponders the future of one particular jumble of letters, RFID, which stands for radio frequency identification. RFID technology, first used in World War II, enables items to transmit information about themselves. RFID tags come in two varieties, active and passive (passive technology was developed in the late 1970s in the United States). Active tags operate off batteries; passive tags are powered by magnetic fields produced by readers. Either type can be slapped on all sorts of products, from a bar of soap to a beer keg to a 10-ton shipping container. And unlike traditional bar codes, which rely on line-of-sight technology, radio frequency tags broadcast signals that are captured by electronic readers. The readers, in turn, enter the data into a company's inventory-management system. "There are substantive challenges ahead," says Ellis, "but RFID is fundamentally better than bar codes."
RFID has been garnering much attention in supply-chain circles of late, mostly spurred by executives at retailing giant Wal-Mart Stores, who have made no secret of their affection for the technology. In discussing the promise of RFID-
enabled supply chains during an interview with CFO last fall (see "Radio Flier," November 2003), Wal-Mart spokesman Tom Williams described RFID as "akin to moving from tapping out telegraph signals to moving to the Internet."
Others agree. In a recent survey of corporate executives conducted by consulting and managed-services firm BearingPoint, nearly 45 percent of the respondents described RFID as a revolutionary technology that will have widespread impact. Spending on that revolutionary technology — currently pegged at $1 billion — will triple by 2007, predicts the Wireless Data Research Group. Technology vendors, which know a good thing when they see it, are already playing up the new technology. "RFID will enable a whole new generation of optimization of supply chains," says David Hushbeck, director of DirectTech, a unit of Menlo Worldwide. "This is a fork in the road."
Bad News For Beverages
Unilever futurist Ellis, who is based in Trumbull, Connecticut, sees the vast potential of RFID. "It could be a paradigm shift," he notes. "Managing space will be a much different proposition if you can ID cases from 20 feet away."
Back in the present, though, many companies are having trouble identifying RFID tags from 20 inches away. For the reality is that RFID — this revolution in supply-chain management — is still very much a work in progress. While many vendors do seem to be gravitating toward a single operating standard, promulgated by business/academia partnership EPCGlobal Inc., RFID products remain expensive. Information-technology consulting firm AMR Research estimates that a typical consumer-products company will have to spend upward of $20 million to roll out an RFID system (see chart below).
|The Price of Tagging|
What a consumer products company shipping 50 million cases a year might spend for RFID implementation (in millions).
|Tags and readers||$5-$10|
|Changes to existing supply-chain applications||$3-$5|
|Data storage and analytics||$2-$3|
|Source: AMR Research|
More problems: passive RFID tags don't work overly well with certain items, including those that are filled with liquids or are shipped inside metal (a fact that can't thrill beverage-company executives). The scrambling effect of those two substances can wreak havoc on read rates. Unilever, for instance, is currently conducting several RFID pilots to see what kind of rates it can expect from the technology. To date, the percentages have been in the 50 to 70 range — a dismal showing. "We've fiddled with the stuff," says Ellis. "But so far, the experience has been a little underwhelming."
Indeed, executives currently contemplating the business case for RFID rollouts — and 86 percent of the respondents in an Accenture survey said they're doing just that — face some harsh realities. Lyle Ginsburg, a managing partner at the consulting and outsourcing firm, has been fielding an increasing number of inquiries from corporate executives about RFID technology for the past several months. Ginsburg says he likes to give prospective clients the bad news up front. "We tell them the costs are too high, performance isn't there yet, the business case is bad, China is threatening its own standard, and the public perceives tags as spy chips — you know, the mark of the devil."
Don't Cross the Streams
And the good news? A whole lot of companies have until January 1 to get their RFID systems up and running. The deadline was set by management at Wal-Mart, which is eager to see its top 100 suppliers tag cases and pallets with RFID chips. Other large procurers, including the Department of Defense, grocery-store operator Albertsons, and retailer Target, have also set RFID-compliance deadlines for suppliers.
It's unclear whether all of Wal-Mart's top 100 suppliers will meet the New Year's deadline, or how many product lines each supplier will initially tag. Although DirectTech's Hushbeck claims that "RFID technology is further along than people think," pilot programs are revealing all kinds of unexpected problems. Delta Air Lines, for one, has conducted two extensive RFID tests and has managed to increase the accuracy of read rates on tagged luggage to a percent in the high 90s. But to get to that level, says Delta spokesman Reid Davis, the airline's managers had to rethink how ground crews stuff baggage into unit load devices (the big metal containers that go into the belly of airplanes). In Delta's early tests, RFID-tagged bags that touched the sides of those metal containers did not always read properly.
Warehouse projects have turned up similar bugaboos. Early adopters report that large metal beams — not exactly uncommon in warehouses — can reduce the accuracy of tags. So, too, can white noise, particularly the white noise produced by moving nylon conveyor belts. And pallets or cases that don't register with a reader can throw a spanner into the works, necessitating human intervention. Says Allyson Fryhoff, vice president of sensor-based services at Oracle Corp.: "Our projects have required a lot more from warehouse personnel than anyone would have liked."
That's a worry, especially since most large corporations have already spent millions optimizing their supply chains and inventory systems. For her part, Fryhoff thinks some companies will have no choice but to change their processes to get RFID-tagged products out the right door. Such reengineering will likely necessitate a redeploying of existing inventory machinery like bar-code readers and optical scanners. Notes Doug Chaney, president of DC Logistics, based in the Dallas area: "Trying to set up an RFID operation in an efficient logistics center can be very disruptive."
Goal: A Nickel a Tag
Trying to budget for an RFID rollout could prove to be equally disruptive. While storage and analysis of the massive amounts of data produced by the tags will be expensive, tags and readers will be the biggest money drainers. Financially, those tags make sense for makers of higher-priced items — things like DVDs and midticket electronic devices. Hewlett-Packard already affixes RFID tags to its printers' boxes. And expensive active tags, which can broadcast more data over greater distances than passive tags, are ideal for high-end items — think plasma television sets — and goods that get stolen a lot.
Down the food chain, however, the average consumer-goods company will likely spend $5 million to $10 million on passive tags and readers — and that eye-popping bill covers only the tagging of cases and pallets, not individual items. "Lots of people say RFID will replace bar codes," says Deepak Shetty, program manager for RFID and security at consulting firm Frost & Sullivan. "But it won't reach the price point of bar codes over the next five to seven years."
It's not even close right now. Currently, top-tier customers are paying 25 to 30 cents for individual tags, points out AMR research director Kara Romanow, while second-tier customers are paying as much as 75 cents a tag. On the surface, a quarter a tag seems pretty cheap, until you realize that some companies will need to purchase 100 million disposable tags a year. And 25 cents is a far cry from the 5 cent tag that RFID backers predict will one day be commonplace. Says Ellis: "At 25 cents a tag, it's hard to imagine too many categories where there will be a business case to be found."
Others echo that sentiment. In the BearingPoint survey, about 40 percent of respondents said tags will have to cost 5 cents or less at the item level before they invest in RFID technology. For top suppliers in low-margin businesses, the cumulative price of tags could eat away at operating profits. Many will be hard-pressed to offset the cumulative costs of passive tags. Says Bo-Inge Stensson, director, group supply chain innovation, at U.K. packaging giant Rexam: "Consumer packaging is a very lean business. The price of tags must be cost-effective."
Leaving Las Vegas
Cost-effective is a ways off, but buying in bulk could help some operations lower the cost of tags. Managers at McCarran International Airport in Las Vegas, for example, recently signed a contract to purchase 25 million RFID tags from manufacturer Matrics Inc. during the next five years. Samuel Ingalls, assistant director for information systems at McCarran, says the airport's managers not only locked in the price but also negotiated an adjustable rate that's based on market quotes. "We can reset the price every six months," he explains.
The RFID rollout at McCarran highlights the promise — and the pitfalls — of RFID technology. Currently, airport personnel rely on optical scanners to tag and identify the 65,000 pieces of luggage that move through the airport each day. The typical read rate of the scanners, 90 percent or so, means that 6,500 bags could potentially be mishandled on a daily basis. The monetary cost of reuniting a bag with a passenger is steep, too: upward of $150. The PR cost can be even higher. "The airport is a visitor's first and last look at Las Vegas," notes Ingalls. "We don't want this to be the last memory of Las Vegas."
Airport officials began testing RFID technology last year. The goal is to ratchet up the accuracy rate of readers to 99.8 percent. That reliability doesn't come easy, though. Readers must be able to distinguish individual pieces of luggage, but in early testing, RFID sensors at McCarran tended to clump bags together. IT managers eventually fixed the glitch by placing reflective curtains on baggage conveyor belts. But overcoming the niggling problem took ingenuity, and much trial and error. Says Ingalls: "We purchased every piece of luggage the Salvation Army had."
The RFID system at McCarran, which includes new baggage hubs, screening devices, tags, and conveyors, will cost approximately $125 million. Much of the bill will be paid by the federal Transportation Security Administration. Managers at private-sector companies, though, don't have the luxury of government funding. The sheer cost of RFID rollouts has some executives — particularly those with demands from large business partners — looking just to meet the minimum requirements mandated by those customers.
Hence, many companies are embracing a slap-and-ship strategy. Others are offloading the problem to third-party logistics outsourcers, such as DC Logistics. Still others hope packaging providers will step in and solve the problem.
The catch in taking such an approach: offloading RFID projects is all cost and little payoff. "When someone asks us to look at the business case just on the compliance side," says Accenture's Ginsburg, "it won't look good."
Those Darned Laws of Physics
Not surprisingly, some corporate managers are looking beyond small test rollouts and stripped-down projects. They see RFID technology as a quantum leap in aligning the moving parts of their businesses.
Executives at Delta, for instance, figure their baggage-tracking system, slated to be fully operational by 2007, will cost the airline in the range of $15 million to $25 million. But Davis says the project will quickly pay for itself, and should throw off savings for years to come. For managers at manufacturing companies, however, wringing cost savings from RFID will be trickier. Businesses are already awash in a flood of data collected from various inventory, supply-chain, and back-end systems. But product data culled from RFID readers will make that flood look like a spritz from a cheap atomizer. "RFID will create billions and billions of bits of data," says Rexam's Stensson. "How do you use this in the best way?"
How indeed? Consultants say the real value in RFID technology will come when companies tie inventory data into back-end systems, then reengineer entire business processes. Once that's accomplished, says Ginsburg, "companies will have true visibility."
It's a nice picture. In the meantime, supply-chain managers are still trying to figure out how to get radio waves to play well with liquid and metal. "Changing the laws of physics," says Romanow, "would really push this thing along."
John Goff is technology editor of CFO.
Radio's Big Comeback?
Problems still lie ahead in adopting workable RFID systems.
The Glitch: Tag prices too high. Why It Persists: Quotes have yet to come down from the 25-to-35-cent range for high-volume tags, and from the 55-to-75-cent range for smaller quantities — too high to be absorbed into a consumer product's item cost.
The Glitch: Tag performance flaws. Why It Persists: Between 10% and 12% of tags are dead on arrival, and even those that pass initial inspections are readable only between 80% and 90% of the time — often because of damage in shipment.
The Glitch: Laws of physics. Why It Persists: Radio frequency doesn't pass through metal or liquid easily, and most pilot projects have evaded that issue. While Gillette and Hewlett-Packard have had some success, and Wal-Mart has said changing tags and antenna sizes will help, outside experts aren't so sure.
The Glitch: No standard. Why It Persists:Two standards organizations, EPCGlobal and the International Standards Organization, are working on it — and companies are waiting for a guideline that won't require too onerous an upgrade.
The Glitch: No network. Why It Persists: Two standards organizations, EPCGlobal and the International Standards Organization, are working on it — and companies are waiting for a guideline that won't require too onerous an upgrade.
The Glitch: No network. Why It Persists: VeriSign was selected as operator of the ePC Network directory, delivering on the Object Naming Service data-sharing vision — but it's still only at the pilot stage, with little industry adoption.
Source: AMR Research