The field deployment of RFID tags has been accelerated by last September’s EPC standards announcement by EPCglobal and Wal-Mart’s tagging mandates. But the EPC is important beyond its use in RFID tags: the “Internet of Things” can and should include intangible things. And the information architecture intended to make an RFID tagging scheme practical will have dismantled one of the major obstacles to product information exchange on the Internet.
It might be healthy for the industry to sharpen the distinctions between three elements: tags, the namespace, and the naming architecture. Almost all of the industry’s focus since the creation of EPCglobal last fall has been on tags. As intangible as the RFID phenomenon is, as an invisible, radio-frequency handshake, tags are something one can appreciate– they have a cost and physical dimensions. But mandates notwithstanding, RFID adoption is likely to come in fits and starts. It may be that some of the physical challenges, e.g., reading tags in cluttered and hostile environments, will render RFID unworkable (or, more importantly, unprofitable) in some of the areas where it has been proposed.
In many ways, the real magic of RFID is in the infrastructure, where a tag’s code can be translated into compact and convenient pointers to sources for information, and the Object Naming Service (ONS) will knit together a richer web of product information.
The EPC was intended to readily encapsulate all of the major legacy product codes and to be large enough to facilitate new applications. There is no reason why one could not assign codes to goods, or even services. An EPC code could be applied to almost anything that one might quantize in saleable units, even if it will never bear an RFID tag.
As an example, one might apply EPCs to airline seats. Like any other perishable commodity, there’s a certain fixed number to sell; the “manufacturers” (i.e., the airlines) could be assigned manager numbers, and themselves parcel out EPCs for every seat-segment they offer.
Such a thing would be unlikely to happen any time soon–the decades-old computerized reservation systems and networks work, and do what they need to do–but one could imagine some movement toward adoption of the EPC as a proxy, especially for new applications. A compact, 12-byte alias for “things with ugly names” (like a seat known today by a combination of airline code, pair of airport trigraphs, etc.), serving as a pointer to more complex, structured content, could be highly useful.
Information for Many Parties
The greatest benefits are likely to come from applications where many, many parties touch uniquely identified “goods” and would benefit more by common identifiers. One could imagine EPCs used to identify music tracks, a virtual good widely exchanged on the Internet, but derived from physical goods (music albums) with existing unique identifiers (UPCs and EANs). While the previous generation of product codes is fairly constrained (100,000 unique products, for a UPC), a music publisher would have orders of magnitude more space to play with, and assigning unique IDs to each and every music track from every published album would fit handily within the EPC’s namespace.
The ONS itself should also have an enormous impact on Internet commerce.
Before the end of the dot-com bubble, some half a billion dollars in venture capital investments were poured into companies creating “codes-to-content” services. These mapped product codes (UPC and EAN bar codes and proprietary extensions) to information and services, such as lowest-cost retailers, on the Web. They included Barpoint, IQorder, Airclic, and DigitalConvergence. The last was famous for distributing several million of its CueCat bar code readers to subscribers of Wired, Forbes, and other magazines–one can still find numerous CueCats on eBay though the company itself is now long defunct.
Each of these code-to-content services ventures tried to construct their own, proprietary pools of content and vendor links. Each faced serious (and ultimately, fatal) network effects issues. The need for end users to want to scan print bar codes was a big hurdle, though it’s increasingly likely that consumers will possess sufficiently capable devices (e.g., cell phone cameras capable of reading 2-D codes) and will be connected… this may have been a case of being a few years too early. A greater challenge was to accumulate a sufficiently large and interesting enough collection of information to satisfy consumer curiosity, ideally through having the actual owners of product informational content (i.e., manufacturers and publishers) provide it. None of this actually happened sufficiently cheaply or easily.
Supply Chain Success
The failed codes-to-content dot-coms were reminiscent of a previous generation of more successful companies, mapping codes to content for supply chain (and not end-consumer) purposes. QRS (public, now being acquired by JDA Software for some $100M) and Global eXchange Services (private, formerly a subsidiary of General Electric, and acquired in 2002 by Francisco Partners for some $800M) both serve as clearinghouses between goods manufacturers and others in the supply chain who handle those goods, for product information interchange. Retailers have their suppliers provide product information to one or both of these intermediaries, as well as to the UCC’s own UCCnet, reducing a messy, many-to-many network of relationships to a simpler exchange via these industry information hubs.
The future, fleshed-out ONS and its dependent services may eclipse this previous generation of supply-chain hub services. The ONS-based architecture for the EPC will encourage manufacturers to make their product information more accessible, and will make knowing a product code–whether an EPC, or the legacy codes they’ll encapsulate–necessary and sufficient to fetching product information. While they’ve served an important purpose as intermediaries, the previous generation of “information silos” aggregating product information could be supplanted by the federated, network architecture promised by the ONS, with myriad applications harvesting content from myriad sources far more nimbly. And that might very well revive an interest in, and provide a much more viable platform for, end consumer code-to-content plays.
Ross Stapleton-Gray, Ph.D. provides information and policy consulting services with Stapleton-Gray & Associates