>> Home > > News > > In the News

Building a Better Supply Chain

February 1, 2002

With these characters, scientists at the Massachusetts Institute of Technology (MIT) have figured out that it's mathematically possible to give a unique ID number to every individual atom on the surface of the earth.

Nobody wants to do that, but the folks at the Auto-ID Center have something almost as ambitious in mind: an open-standard, automated supply chain where all products are identified by a unique electronic product code carried on antenna-mounted computer chips the size of dust specks. Whether it's called the next-generation barcode or smart identification, the new technology could cut billions of dollars from supply chains all over the world.

It's also the kind of pie-in-the-sky vision that can be easily dismissed as science fiction, if it weren't for the down-to-earth companies that have invested in the vision. The Auto-ID Center is an industry-funded research program ($300,000 per sponsor) based at MIT. Among the investors are packaged-goods heavyweights Procter & Gamble, Johnson & Johnson and Gillette. Wal-Mart Stores is on board. The Uniform Code Council has been involved from the beginning.

The program isn't close to achieving its vision, but there are some advances. Right now in Tulsa, Okla., the project is testing a limited pallet-identification program in and around a Sam's Club.

According to Brooke Peterson, associate director of the Auto-ID Center, the test has so far shown that the system works at a basic level. "In phase one, everything went beep that was supposed to go beep," she says. One exception: A truck broke down delivering some of the tagged pallets.

While moving pallets of paper towels, and at a more advanced phase, individual cases of paper towels, the field test makes sponsors quicker-picker-uppers of information.

At a basic level, participants in the test are taking the electronic product code (ePC) for a spin. Similar to the 24-digit code mentioned above, the coded RF signal will be picked up by readers tracking the shipments as they move across the supply chain. At a deeper level, the ePC is being converted to a Web or server address containing information about the pallet by way of an Object Name Service, similar to the more popular Domain Name Service. Computers will be directed to a URL address containing information about the pallet. The physical object associated with each ePC also is being described through physical markup language (PML), based on eXtensible Markup Language, allowing multiple computers to share information about the object in a common format.

Phase one tracks pallets of merchandise, with the intent to debug the ONS and PML software developed by the Auto-ID Center at MIT. The second phase of the test will advance from pallets of merchandise to individual cases and test the system's ability to handle greater volumes of inventory and information.

Wal-Mart declined to comment to Chain Store Age about the progress of the test, which will last at least until May.

The current test identifies pallets and cases of goods--the sorts of things that are identified with a linear barcode, the UPC/EAN-128 code, at most warehouses today. But can a technology that marries computer chips and miniature RF antennae ever replace the standard barcodes printed on everything from toothpaste to a loaf of bread?

Cost remains a barrier: The current price of RFID tags ranges from 25 cents to a dollar, according to a Texas Instruments scientist. At that price, it's economically feasible to track cases or pallets of merchandise, but not an individual can of Coke. To really achieve the long-term goal of the Auto-ID Center, the cost of creating and embedding a tag for everything in the store has to shrink dramatically.

In a speech given at the Low Cost, Smart Label Conference in Cambridge, England, Auto-ID Center director Kevin Ashton stated, "I'll be frank. We have no idea how to build a wirelessly networked computer--because that's exactly what an RFID tag is--for a penny or less, let alone for free. At least not yet."

That's not to mention the opposition from libertarians who oppose introducing technology that could conceivably track their every move. Peterson, however, says customers will have a choice to kill the tag at the point of sale.

Chip-making techniques utilizing acid instead of diamond saws are an exciting development to reduce those costs. Ashton thinks a five-cent tag is feasible, and that a prototype should be available sometime next year.

Ashton's big-picture message: The world is on the brink of an identification revolution. The barcode will continue to be effective, but it has limitations that radio frequency does not have.

"The barcode's biggest limitation is that it is an optical technology, requiring a clear line of sight," he says. "Barcodes help us to identify things. They do not identify things on their own."

The Auto-ID Center, Ashton says, was formed to create a "single, open-standard, very low-cost system that would allow computers to identify physical objects."

How soon before something really happens? At the end of 2003 or early 2004, smart technology could be adopted for real, he says, most likely in supply-chain uses where pallets, rather than individual items, would be tagged.

He also makes the educated guess that by 2004, we will see the supply of RFID tags in the low billions and the realistic, profitable tag price of 5 cents. Billions, however, are a drop in the bucket. And even that development is contingent on several variables.

IBM consulting marketing manager Dan Hopping, who has been working on RFID applications for five years, believes this is going to happen. The only question is, how fast? But he cautions that building the chip is the easy part, the hard part is the antennae for the individual chips and the infrastructure throughout the supply chain.

SIDEBAR:

The Gap has shown that radio frequency identification (RFID) on individual products is a matter of science fact, not science fiction.

The retailer completed a three-month test in Atlanta during which denim apparel carried RFID tags to help the chain track individual items as they passed from the factory to store shelves. The result: higher sales for RFID-tagged denim apparel compared to stores in the area not using RFID technology.

The Gap is not saying how it intends to proceed with RFID tags, but Dallas-based Texas Instruments, which provided the 2.25-sq.-in. tags, says retailers don't have to wait to experience the benefits from RFID technology.

During the test, store employees were spared the chore of manually reading the barcodes of individual items as the denim products entered the store. Instead, a tunnel reader in the back of the store automatically received the signal from the individual merchandise as it arrived off the truck. Readers on the shelves and at the checkout also tracked the items automatically. Replacing the line-of-sight dependent bar code reduces human error, according to Gap and Texas Instruments.

Gap's test differs from that of the Auto-ID Center in several areas. Chief among them is the closed vs. open system. Gap's was closed.

"You will see RFID technology appear first in applications we call closed," says Tres Wiley, strategy manager for emerging markets, Texas Instruments RFID Systems. "That's where one party owns the point location where the tag is put on and accrues the benefit of reading the tag throughout the life of the item."

Wiley says Texas Instruments was approached to participate in the open-system RFID project based at MIT, but declined. "We decided that while their long-range objectives were right on target and highly appropriate, there are lots of applications that can be solved with today's technology."