Revisiting RFID -- Again


In writing about Gartner's latest HypeCycle report last week, I noted somewhat snarkily that RFID had been "mired in the Trough of Disillusionment for an awfully long time." Gartner projects that RFID will become mainstream within a decade (in the company of such technologies as surface computers, 3-D printing, virtual assistants and context-delivery architecture).


Seems like I heard that, oh I don't know, about 10 years ago.


RFID never seemed to live up to its ample potential. Or maybe people just expected too much of it. The tech industry does love overheated marketing, but perhaps no technology in recent memory was hyped as vigorously as RFID. Check out this Scientific American article from 2004 that bills RFID as "the key to automating everything." It describes a society in which RFID chips and readers are nearly ubiquitous and will transmit information to computers that then do everything from turning on your shower -- adjusted to the ideal temperature -- to telling you what you need to buy at the grocery store.


Reality has been more mundane -- and not always successful. A strategy+business article points out that expectations are being scaled back. Instead of being perceived as the cure to all supply-chain ills, RFID is being employed in smaller-scale, closed-loop applications, a few of which are mentioned.


For every big success, such as Land Rover using RFID to track vehicles at its factories, a project that yielded a full return on investment within nine months, there's a deployment with more muted results. For example, Gillette uses RFID to ensure its products get to store shelves faster and expects a 25 percent ROI by 2016. Achieving a 25 percent ROI within a decade doesn't sound all that compelling to me. The article also briefly touches upon two of RFID's big sticking points: its reliability, which is still far from rock solid, and the price of tags, which makes the technology too expensive for many of its proposed applications.


While retail supply chains were initially positioned as RFID's killer app, the technology is proving more immediately useful in non-retail environments. Use of RFID at hospitals has tripled over the past three years, according to Spyglass Consulting Group. And according to a Network World article, hospitals are moving from using RFID for simple asset tracking to using it to streamline work flows and to improve health care processes. Some respondents to a Spyglass survey link RFID data to process-improvement programs such as Six Sigma.


Still, RFID tends to be used mostly to solve department-level problems rather than to address broader hospital issues. Moving beyond that is "not for the faint of heart," says Greg Malkary, Spyglass Consulting Group's founder and managing director. He says:

You need a full embrace of these technologies, better partnership with vendors and a new level of middleware that can glue everything together. Right now, this is done with custom code to bring all this together. It's a big investment.

A fast-growing use of RFID is to monitor IT assets, says Aberdeen Group analyst Michael Dortch. It's due to improvements in tags, which have been tweaked to better transmit signals in environments with a lot of metal, such as data centers, according to an InfoWorld article. Tags are also now small enough to be placed on individual blades in server racks, where data can be read without disturbing the racks.


IBM has an application called Services Assets, which employs RFID to track IT gear as well as other stuff like office furniture. The real value comes from integrating the RFID data with full-blown asset management tools from CA, IBM Tivoli, and Microsoft that track information such as what software is on the hard drive, what chip sets are on the server, and which versions of operating systems are being used.