More
Without such a kick-start, RFID was backed into a classic chicken-egg impasse: it needed a broader user base to improve technology and drive down cost but until the technology got better and cheaper there was no incentive for users to sign on. That all changed with Wal-Mart's mandate and the confirmation that eight global manufacturers-Gillette, Johnson & Johnson, Kimberly-Clark, Kraft Foods, Nestle Purina, Procter & Gamble and Unilever-would participate. In one stroke, there was critical mass.
The pilot initiative was hardly undertaken in a vacuum nor was it unwelcome. EPC itself began in earnest as a collaboration between P&G and Gillette with the Massachusetts Institute of Technology to form the Auto ID lab in 1999 (subsequently 100 companies and five other academic institutions would join the federation). Its mission was grandiose: to generate the intellectual fire-power to develop an open standard architecture infrastructure that would "make it possible for computers to identify any object anywhere in the world."
An industry standards body, EPCglobal, evolved from that beginning, established to 'increase supply chain effectiveness' across all industry sectors. Blue-chip chief supply chain and chief technology officers sit on the board, along with the Assistant Deputy Under Secretary of Defense for Supply Chain Integration and Linda Dillman, Chief Information Officer at Wal-Mart. It was out of this context that Wal-Mart's pilot mandate was issued.
Wal-Mart's pilot, which involved select Supercenters and a Distribution Center in the Dallas-Fort Worth region, is the biggest real world trial for RFID. As such, it is on the radar screen. And, the question everybody in the supply chain asks is "how's it going?"
Wal-Mart's appraisal
According to Wal-Mart, the pilot is proceeding on schedule although some observers note that "100 percent compliance" has become more of a moving target than the original target of January 1. The company intends to move forward with plans in place to have RFID readers installed in as many as 12 distribution centers and 600 stores by the end of the year."We're using RFID in our stores in the Dallas area, providing visibility we never had before," Simon Langford, Wal-Mart's manager of global RFID strategy, told Wal-Mart shareholders in June. At last count, well over 90 percent of singulated cases (items not stacked on a pallet) were being read as they came into the store; palleted cases remain more problematic, with a read rate of only 60 percent.
A key discovery during the Texas pilot has been the realization that RFID readers enable management to know the whereabouts of the product in the store. Wal-Mart has installed readers at the door to the sales floor (without RFID, they couldn't tell if the product had moved out of the back room). "RFID gives us that window," says Langford. (Target Corp., which is conducting a similar if less publicized RFID pilot, has let it be known that their primary thrust is aimed at knowing what's in the back room).
Building on that window, the company is looking to RFID data to help determine re-stocking strategy. During peak seasons, employees are able to fill only one of twelve out-of-stock situations on the floor. Wal-Mart is looking toward RFID data to make it easier to locate merchandise in the relatively small back room and prioritize what gets re-stocked first. Facilitating 'pick-lists' exemplifies the multiple possibilities RFID affords. "It's how you use the data," Langford notes in observing how fundamental business processes may be changed.
Of particular value thus far has been a more precise view of how long it takes a product to move through the Wal-Mart supply chain. One surprising note thus far is that many suppliers are discovering that the entire process takes a day longer than they expected, an insight which is likely to affect future replenishment processes.
Meanwhile at its Bentonville, Ark. testing facility, Wal-Mart continues to push the envelope on RFID possibilities. One project entails installing readers on forklifts to enable maximum scanning capability. Another involves a new type of tag that uses ultra-high frequency.
Summing up Wal-Mart's take on RFID thus far, Chief Information Officer Linda Dillman told the National Retail Federation earlier this year, "We don't yet know the size of the impact, but we know it's there in a big way."
The suppliers' point of view
The first successful read in the entire Wal-Mart trial was a case of Scott paper towels, a fact relayed to Mike O'Shea, Kimberly-Clark's Director of AutoID/RFID Strategies, while on vacation ("we toasted to the success with a mango daiquiri, which is now the 'official drink' of RRID"). O'Shea and his team had been preparing for RFID several years in advance of the pilot: "So far," he says, "there are no surprises."But no surprises do not necessarily mean no problems. Those problems fall into two general categories. One is technical. "We're dealing with the law of physics here," O'Shea notes. "The environment has an influence on how radio waves perform. Concrete and steel can influence them. Likewise, humidity can get absorbed into packaging materials." The second class of problem involves the abundance of raw information itself. "How will you take all the data you're going to get and integrate it into your business processes?"
What has K-C, with two facilities currently outfitted with RFID (they are participating in RFID projects with Target and two major European retailers) and two more scheduled to be on line by 2006, learned so far? "We've learned that RFID is not just another form of 'plug-and-play' technology. You have to develop your own intellectual capital." Acquiring this intellectual capital is a staged process. "Get engaged in some way and on a scale that makes sense to you," O'Shea advises. "Don't give integrators 'the key to the family car.' They might know the physics and tech issues but you have to learn what you want them to do as it relates to your own business."
"Everybody's going through a lot of pain now," he admits, "before we get to the end state." That end state is where the technology can be employed economically enough to validate the business models upon which it is predicated.
One company whose RFID business model is unsustainable in the present context is Procter & Gamble. There's no question at P&G whether AutoID will figure prominently in the future. Like K-C, they have been working on it for a long time (P&G was a co-founder of the M.I.T. AutoID lab). "RFID is great technology," says company spokesperson Jeannie Tharrington. "We've learned that the technology is viable. It works. But what we need is standardization, there has to be industry-wide standards."
The cost remains higher that P&G originally thought. The original business case budgeted applied tags to cost less than a nickel a case or a palette. Currently, the application cost for homogenous pallets, in-line applicators can be used to tag cases automatically, is in the neighborhood of 10 cents per case. However for product categories where case picking operations are required to satisfy mixed pallet requirements, application costs can be as high as 50 cents a case.
Where are the glitches? "We had thought volume would be higher by now with more companies would be tagging more products," Ms. Tharrington notes. "We also need a tagging technology breakthrough."
The Wal-Mart pilot has provided valuable refinement to how P&G regards the process of Auto ID. "First," she notes, "we've learned that EPC technology will not allow us to scan all the cases in a pallet, especially when the product in the pallet has liquid or metal properties. We've also learned that it is critical to develop a software solution to record the case EPC information before a pallet is completed. And, we realize that portal readers on every door seem to be an unrealistic expectation."
P&G agrees with K-C on the need for more manageable approaches to organizing and messaging the information collected. "Data Management will be a huge issue," says Ms. Tharrington. "You can get all this great info in real time and automated but what will we be able to do with all of it? How will it enable us to do things differently and better? Until we dive in and figure out what the software looks like and the real applications, it's still a whole other ballgame."
Hewlett-Packard finds itself in a special position with regard to the Wal-Mart test, serving both as a vendor of RFID consulting services as well as one of its suppliers. With 26 facilities using RFID in its own supply chain internally, HP was no 'ordinary' participant.
"From the start we've been really proactive about RFID," says HP spokesperson Dayna Fried. "Groups use RFID to track manufacturing. It's used to track cassettes of wafers in semiconductor fabs. We're integrating it with sensors to build intelligent monitor systems. We anticipate we will spend conservatively $150 million in research for our own supply chains."
Moreover, unlike companies such as Kimberly Clark or P&G, which sell low-value products tagged by the pallet, HP's high value products are individually case tagged. "We're not tagging everything we manufacture for Wal-Mart," explains Ms. Fried. "We ship 65 different consumer SKUs to Wal-Mart and 40 to Sam's Club. Of those we're RFID -tagging 60 to 75 percent. A year ago, we were tagging 4 product SKUs." In addition, HP was able to reduce the pallet build speed of some products from 90 seconds to 11 ("a stunning process").
HP didn't have to adjust its processes to adapt to Wal-Mart's RFID directive. "It was easy for us as a tech company but probably harder for others," Ms. Fried notes. Even so, there were hurdles to overcome. "Read rates were a problem in internal tests, especially with ink jets because of the liquid and metal (the solution was to rotate the pallet at the scanner which then yielded 98 percent read rates). We learned a lot about the care and handling of tags." There were insights about necessary infrastructure ("you need a high availability of local server back-up for storing information"). Tags remain a cost factor ("We buy in bulk at between 15 and 50 cents apiece. We anticipate the cost going down with greater demand").
The bottom line on RFID? "We're waiting for others to give the green light to beef up what we're doing. But they need to have the technology in place to accept the product and to store all the information. It's all new and going to ramp up a little slowly."
Net-net
If there are vendors explicitly unhappy with the Wal-Mart pilot, they are keeping a low profile. What has struck participants is the spirit of collaboration amongst a wide front of enterprises with a stake in improved supply chain visibility. This is particularly true looking ahead to the next stage of RFID development, Generation 2, the standards of which are now being forged. "We've got all parties-competitors and members of the value chain-working together to create interfaces that can be used globally," says Kimberly Clark's O'Shea."Everyone knows where we need to go. How fast we get there will vary from company to company. The vendor community is working hard to get the product out."
Meanwhile industry analysts anticipate a steep upward curve of RFID sales. Within two years sales of readers, tags, software, and services are expected to pass $5 billion (up from $1 billion in 2003). "We're starting to see a lot of non-supply-chain applications worldwide," says analyst Reik Read of the investment firm Robert W. Baird & Co. "Exploration efforts from Wal-Mart and the Department of Defense are starting to pan out in the form of applications you would have never expected."
Sidebar
EPCglobal: Pulling the Strings of RFID
By Marlo Brooke
If all the world's a stage, RFID can be likened to an actor whose movements are dictated from behind the scenes by EPCglobal, Inc., the inter-industry organization whose mission is to revolutionize supply chain efficiency.This not-for-profit entity has effectively established itself as the global standards organization for RFID in supply chain. As Christi Gallagher, spokesperson for Wal-Mart points out, EPCglobal is "the only organization we endorse in the RFID space."
Endorsed by GS1 US (formerly Uniform Code Counsel), and GS1 (formerly EAN International), which represent well over 1 million companies worldwide, EPCglobal spans continents. And industriesâ-from automotive, to healthcare and life sciences, to government and defense. Its Board of Governors include luminaries like Linda Dillard, EVP and CIO of Wal-Mart, Alan Estevez from the Offices of the Secretary of Defense, and top executives from DHL, Hewlett-Packard, Cisco and Procter & Gamble.
EPCglobal originated as the Auto-ID Center in 1999, an academic research project headquartered at Massachusetts Institute of Technology (MIT), with labs at five leading research universities worldwide. Its objective was to develop an open architecture for a global supply chain network, which it accomplished near the end of 2003. At which point it changed its name. As EPCglobal. Inc. it evangelizes a vision throughout the world while RFID research continues behind the scenes through another entity known as Auto-ID Labs, a close cousin of Auto-ID Center and EPCglobal based in 7 universities and funded partly by GS1 and GS1 US.
"In simple terms," states Jim Williams, senior consultant at Avatar Partners, "the objective of EPCglobal is to reduce costs in supply chains that are exponentially crippling global business today." EPCglobal addresses this daunting task by establishing and supporting the global adoption of the EPCglobal Network, a communication channel to connect servers containing information related to items identified through a "peer-to-peer" Internet-based network (with EPCglobal acting as security monitor and trafficker).
EPCglobal Network has five major components, all of which are designed to help trading partners collectively make better decisions and increase automation. These components are:
- Serialized Electronic Product Code (EPC), through which any given item can be uniquely traced throughout the supply chain
- Identification System-identifies the item, including the physical RFID tags or containers, and RF readers
- EPC Middleware-provides backend database integration for intelligent decision-making and automation, and eventual communication to EPC Information Services
- EPC Information Services-the physical location where data collected about that item is housed, whether on an existing ERP or standalone system
- Discovery Services-a suite of services that allows respective trading partners to find data about any item at any point in the supply chain
The vision of EPCglobal goes beyond simply RFID to encompass a totally networked economy in which there is perfect synchronicity between supply and demand, (so that the moment an item is used, purchased or consumed, an automated trigger produces the right and appropriate action to the value chain suppliers all the way back to raw goods procurement).
RFID's role is significant in this Synchronized Value Chain as an ideal technology for essentially computerizing every widget that a company produces. This is a major paradigm shift; up to now, supply chain accounting has been based on processes that are automated at best, manual at worst, and generally not well tracked at detail level. RFID holds out the promise of granular visibility.
Marlo Brooke is president of Avatar Partners (www.avatarpartners.com), a global consulting organization specializing in supply chain best practices. She welcomes your email at mbrooke@avatarpartners.com.
With access to over one million professionals and more than 60 industry-specific publications,
