A chance encounter four years ago between two professors and an industry executive in Cambridge, Mass. may one day have a substantial impact on supply chain efficiencies.

That day, Sanjay Sarma and David L. Brock, professors at the Massachusetts Institute of Technology (MIT), and Kevin Ashton of Procter & Gamble met at an MIT conference. The two professors described their research into low-cost radio-frequency identification (RFID) tags. Ashton noted industry's interest in using such tags in the supply chain. 'It was a meeting of the minds,' recalls Sarma.

As a result, MIT's Auto-ID Center (617-452-2521) was formed. Starting in 1999 with corporate sponsors including P&G, Gillette Co. and the Uniform Code Council, the center immediately launched research into RFID tag development and usage.

Today, Ashton handles the business issues while Sarma and Brock direct research. Meanwhile, more than 48 companies sponsor the research center. Sister labs have also been set up at the Cambridge Institute for Manufacturing in Cambridge, England and at Adelaide University in Adelaide, Australia.

In its nearly four years of existence, the Auto-ID Center has struck out on multiple paths to bring RFID tags into widespread use in the supply chain. These include development of tag technology, protocols, standards and practical applications. The hoped for result is low-cost tags that easily exchange information at multiple points in the supply chain. That would then trim costs and time while making more information more available to more decision makers.

Getting to that point, however, is a complex process. Tags, sometimes called transponders, contain an integrated circuit and an antenna to transmit the information stored on the chip. A reader or interrogator receives the data transmission and relays the information usually to a computer. Unlike bar codes, transponders sending data via radio signal do not need a perfect line of sight with a reader to exchange information. Instead, the information can be exchanged without the reader ever 'seeing' the chip, making RFID a more flexible way to swap information.

When the center got off the ground, Sarma says he and Brock believed that the lack of low-cost tags was hindering widespread deployment of the technology. Today, an individual tag costs in excess of 50 cents, Sarma says. However, if tag cost dropped to a nickel or less, then more companies would be willing to use them. 'We felt that RFID tags could be made less expensively with a change in the systematic approach,' says Sarma.

A systematic approach was required because RFID tags themselves are complex systems. The design of the optimal tag requires a careful coordination of the integrated circuit design, the antenna design, and the manufacturing process itself. A system-level approach was necessary to bring about the introduction of a low-cost tag.

A particular rub was the makeup of the existing tags. Because the tag industry was not taking advantage of advances in communications, it was forced to incorporate excess functionality on the tags, which resulted in higher costs, explains Sarma. He notes that existing RFID tags require extensive memory to encode data. 'If the footprint was shrunk, the chip (inside the tag) could be tinier,' said Sarma. 'If you make the chips smaller, the costs go down.' Lower cost chips could then be churned out by manufacturers in higher volume, driving up demand.

Sarma said he and Brock believe that a three-part solution will fix the problem. For starters, data on the tag could be minimized by keeping information on a back-end network. In other words, the tags would carry only a 'license plate number' or pointer that would identify the marked object. The license plate, in turn, corresponds with item-level information stored in a computer or on an information network, which could even be the Internet.

Also required are new protocols - a set of operating procedures for tags - to interact with readers. Sarma says that the Auto-ID Center has already developed protocols for tags sending radio signals on the 13.56 Megahertz (MHz) band, which is used most widely outside the United States. It has also developed a protocol for 915 MHz, which is used extensively in the United States.

Finally, low-cost tags would require the introduction of new, thinner chips. Sarma says the Auto-ID Center has lined up three silicon manufacturers to produce these new chips later this year. 'By the third quarter of this year, we hope to have this prototype silicon,' Sarma said.

The Auto-ID Center has also begun championing the concept of an electronic product code (ePC). Designed to mark physical objects such as pallets and containers, the ePC provides a license-plate reference to the item-detail information about the object held on an information network.

The ePC works together with a product markup language (PML) and an object naming service (ONS). PML is a new standard 'language' for describing physical objects to the Internet in the same way that hypertext markup language (HTML) is a common language on which most Internet Web sites are based. The ONS tells the computer systems where to find information about any object that carries an ePC code, or smart tag. Moreover, ePC would be compatible with the current universal product code (UPC) and global trade item number (GTIN) identification bar codes. The former is the bar code on U.S. retail items and the latter covers retail for trade items used worldwide. Such compatibility would allow a manufacturer to substitute an ePC tag for a traditional UPC bar code printed on a label. The advantage of the ePC is that they are so-called 'smart' labels, carrying more information than conventional bar codes.

The center has written new software for handling the data routing to support the ePC. Five leading companies - P&G, Gillette, Coca Cola, Wal-Mart and Unilever - have tested the software using current equipment for tags and readers. The companies are now in the first phase of testing pallets with tags. When these pallets arrive at a warehouse, a reader placed on the dock door reads the tags to confirm arrival. In the next phase of the test, companies will use a new generation of readers, Sarma said.

Sarma says that the Auto-ID Center hopes to establish uniform standards for RFID tags used in the supply chain but will leave it to the various transponder manufacturers themselves to produce the ePC tags. 'We want everything to be done with open standards and interfaces,' he says.

The research director is confident that the adoption of ePC tags will revolutionize supply chain practices. Such smart labels will provide for a higher degree of information sharing among supply chain partners. Sarma suggests that the optimization of the supply resulting from the use of RFID tags should make the economy more efficient and improve our standard of living. 'We really want RFID to be a much larger, valuable part of our economy,' he says.

Click this icon to read more about the history of Auto-ID Center.