numbering schemes and standards

This page considers RFID standards and numbering schemes.

It covers -

• introduction
• numbering schemes
• standards and information sharing

     introduction

First generation chips support 40 bit identifiers. It is expected that many applications will rely on an identifier in the range between 64 and 128 bits. In principle use of 96 bits would allow 79 octillian tags, glossed by one observer as enabling every person on earth to have 50 quadrillion tags (presumably with many tags being used on a once-only basis).

From a technical rather than policy perspective there is little difference between "giving every cow in the United States its own unique identification number" and having a unique RFID number for every person in the US, Netherlands or Australia.

     numbering schemes, standards and information sharing

As the preceding page suggested, no single standard numbering scheme for RFIDs has been established. That is unsurprising, given the diversity of applications and stakeholders, and will presumably remain the case in future.

The scheme with the highest profile is the Electronic Product Code (EPC), which traces its origins to consumer product supply chain category/item identification initiatives such as the EAN and UPC/UCC barcode ... a standard way of expressing information about physical entities across individual organisations, industry sectors and nations.

It is thus analogous to the ISBN, ISSN and other schemes discussed in the Metadata profile elsewhere on this site.

The EPC, under the auspices of EPCGlobal, has been promoted as

the next generation of product identification ... a simple, compact 'license plate' that uniquely identifies objects (items, cases, pallets, locations, etc.) in the supply chain. The EPC is built around a basic hierarchical idea that can be used to express a wide variety of different, existing numbering systems, like the EAN.UCC System Keys, UID, VIN, and other numbering systems.

Like many current numbering schemes used in commerce, the EPC is divided into numbers that identify the manufacturer and product type. But, the EPC uses an extra set of digits, a serial number, to identify unique items.

EPC numbers comprise a

• Header, identifying the length, type, structure, version and generation of EPC
• Manager Number, identifying the company or company entity
• Object Class, similar to a stock keeping unit (aka a SKU)
• Serial Number, the specific instance of the Object Class being tagged

and optional supplementary fields, used to properly encode and decode information from different numbering systems into their native (human-readable) forms.

RFIDs are not (and will not) be restricted to EPC numbers. Bioinformatics, library, traffic management and other applications use numbering schemes of varying complexity that are specific to particular organisations or sectors.

Many of those schemes will not be meaningful to unauthorised observers unless there is access to contextual information, eg a unique item identifier may be meaningless on its own but meaningful when related to a bibliographic database that features data about the entity associated with that tag, where the entity is held and who has borrowed the entity.

One concern of consumer advocates is that there will be substantial sharing of customer-related information collected via retail tags, given recognition that major retailers such as WalMart have major information collections are an incentive to engage in sophisticated datamining.

In practice some of those concerns appear misplaced, as major players have tended to be reluctant to share granular (rather than aggregate) information with competitors and third parties.

    next page  (issues)