Radio frequency identification (RFID) has been hailed as a "super-barcode" for retail, enabling goods to be tracked efficiently throughout the supply chain without the need for direct contact or line-of-sight scanning.
Enthusiasts also see great promise for RFID in a host of broader applications, including access control, baggage handling and fraud prevention. But how will that promise be fulfilled in the real world?
As with many technologies aimed at mass take-up, it all boils down to effective standards. Choose the right set of standards for RFID and the gates for mass adoption will open. Make a mistake, and RFID could get caught in a technological cul-de-sac.
The consensus among analysts seems to be that RFID standards are mature enough to support barcode replacement. The standards for more complicated supply chain management capa-bilities will emerge soon. But individual item-level tracking standards, the functionality around which much of the RFID hype revolves, is still some way off.
In fact, there are still basic technological problems to be overcome, let alone standards. How, for example, can products such as those wrapped in metal foils be successfully tagged with chips that might receive interference from metal wrappings?
Knock-on problems also compound the uncertainty around standards. For example, manufacturers of RFID readers tend to be unwilling to invest heavily in devices that may become obsolete. Conversely, users are hesitant to spend as it may lock them into a supplier that goes bust.
"All readers should be field-tested with the brand of RFID tag intended for its use," said Bruce Hudson of Meta Group. "If something is 'EPC compliant', this should not be interpreted as a problem-free guarantee."
Another standards problem is with applications beyond retailing and distribution. Secured access control, anti-theft, container or boxcar tracking, toll taking, asset tracking, baggage handling, livestock tracking and item-fraud protection all have complex inherent functions that are barely touched by RFID.
"The primary problem is that supply chains are unavoidably diverse in nature, comprising multiple systems, applications and operational capabilities - and the RFID chip technology will be expected to interoperate seamlessly across the board," said Alan Lawson, a research analyst with Butler Group.
"A failure at any single point could easily impact activities up- and downstream, and users will not commit to such systems until it has been decisively proven that RFID has an ability to perform in a robust and dependable manner."
RFID forces businesses to address difficult matters, including privacy, collaboration and return on investment.
"It is evident that standards are key for collaboration," said Tony Hart, enterprise applications managing analyst at Datamonitor.
"Collaboration will demand greater use of standard communications and messaging to ensure its success. Global data synchronisation will become fundamental and, for many organisations, this will lead to a review of existing business-to-business infrastructures and e-business strategies."
However, some of the responsibility relating to standards falls on the shoulders of internal IT departments. These include:
Scoping the breadth of an existing infrastructure and application portfolio
Understanding RFID's capabilities and limitations
Setting privacy policies, especially those that address tag deactivation.
"The impact of RFID must be assessed across the entire infrastructure and application portfolio, because RFID will touch all aspects of the enterprise's IT investments," said Enrico Camerinelli of Meta Group.
"For cutting-edge organisations, we believe an RFID 'pilot health check' should be performed before moving on to a production roll-out. This check is to ensure that all product types have been tested and that the design has addressed issues such as scale and velocity."
When it is running, RFID will generate lots more data for companies to handle and this, in turn, will create its own set of problems. A UK trial called cd.id, between Asda, EMI and distributor Handleman, RFID tagged 8,000 CDs. The trial indicated to Rob Salter, managing director of Handleman, that RFID is still "a work in progress". Or, as he put it, how the individual tagging of hundreds of CDs in a single box enhances knowledge of product movement across the supply chain is still far from clear.
"The fear of choosing a potentially wrong standard is preventing many companies from launching pilot RFID projects," said Camerinelli. But he is convinced that the option of doing nothing is a mistake. "The benefits of getting acquainted with the technology and the ability to shape processes around it will offset the risk of choosing the wrong standard," he added.
For example, current production-ready RFID does offer resolutions to some of the standards hurdles, such as tag size, read and write capabilities and radio operating frequencies.
"A rail transportation organisation's use of RFID to track entry and exit of boxcars in a rail yard has spawned a similar design model for distribution centres or stores," said Camerinelli.
"In this model, goods will be placed into an inventory on entry through a gate and removed from an inventory when the goods pass through another gate." Alternatively, the pilots of leading retailers have RFID reading gates placed at the loading-dock areas, the doorway to the store floor, and the doorway to the disposal, as well as in point of sale stations.
Legislative deadlines will also force standards resolution, perhaps faster than market pressures. Forrester senior analyst Charles Homs, reflecting on the impact of the EU's Health and Consumer Protection Directorate that will come into law on 1 January 2005, said, "To address the problem of food traceability, retailers and packaged goods firms should use RFID to meet traceability compliance deadlines.
"Regulations do not specify the use of RFID to comply with food safety regulations, but using them to find goods in distribution centres, stores and trucks will help firms respond within the time limits to any official inquiry."
The IT industry has recognised the potential of RFID. Accenture's chief scientist Glover Ferguson, for example, has long been advocating their inevitable ubiquity. Software suppliers including Microsoft and Oracle have recently launched initiatives, a good sign that the hurdles are coming down. And with support from systems integrators, there is no lack of energy or investment to cure RFID of its current standards headache.
2004 Newcomers will run pilots and test operating environments. There will be some limited deployments
2005 Pressure from the likes of Wal-Mart and the Department of Defense will force the issue of compliance
2006 RFID capabilities will be widely extended by retailers, distributors and suppliers
2007 New issues will come to the fore as RFID adoption forces more complex compliance requirements, such as integration with enterprise applications and across verticals
2008 Standards activity will reach resolution.
Carrier frequency The radio frequency at which an RFID tag operates. Different systems are used to transmit data including Amplitude Shift Keying (ASK), Frequency Shift Keying (FSK) and Phase Shift Keying (PSK)
Data rate The speed at which data is exchanged between a tag and a reader
Error management Standards used to check that data is being exchanged correctly
EPC Electronic product code: the standard for uniquely identifying products
EPCglobal The retailers' organisation with the job of standardising item numbering
ETSI European Telecommunications Standards Institute: the European body
Microsoft Radio Frequency Identification Council The new Microsoft-led group which includes Accenture, GlobeRanger, Intermec Technologies and Provia Software. The council is working on an RFID platform
Passive transponder A tag that only operates when triggered by a reader (active transponders have their own power source and operate universally)
Read/write The ability to update the data a tag carries as well as read it
RFID Radio frequency identification tagging
SRD A short-range device: less than 100mm.
This was first published in July 2004