RFID: Nearly Ready for Prime Time

The wireless tag reader technology may not be mature enough for widespread enterprise use yet. But some users are hoping that 2006 will be the magic year.
Sometimes being an early technology adopter can be a downer. But Bill Miller, CIO for El Paso County in Colorado, plans to use what he has learned from his failed radio frequency identification (RFID) project next year, when he says the technology will be ready for prime time.

''Early adoption is generally a lot of work. We were expecting that,'' says Miller. ''But RFID just isn't sophisticated enough yet. The technology is extremely immature.''

Those are harsh words for the wireless tag reader technology that has garnered backing from the likes of retail powerhouses Wal-Mart, Starbucks and Gillette. IDC, a research firm in Framingham, Mass., predicts the RFID market, including services surrounding implementation, will reach $2 billion by 2008.

But for Miller, the proof is in the $70,000 pilot.

Miller hoped to use the technology, which lets users place detailed information on a wireless tag and then match that information via a reader to a back-end database, to ease his asset management woes. He planned to place RFID tags on the 2,300 desktop systems throughout the county to track the life cycle and location of both hardware and software.

''Computers might show up on the network, but they're not always in the same place they were six months ago,'' Miller says. He has a bootstrap IT operation with only three full-time desktop people. However, he replaces more than 400 systems in a year -- finding it ''less expensive to replace them than to fix them.''

With the RFID tags, his team would be able to do a quick walk-through on a floor of a building and assess the location and age of the desktop equipment by matching the tag with vital information, such as purchase date and licensing or lease agreements, stored in an Oracle inventory control database.

But a few months into the project, which started with the IT group's 200 desktops, Miller found out that the readers were not mature enough for the task. Instead, distance and interference limitations would not even allow for the eight feet of space he needed for rapid reads.

''If you have to get right up to the PC, then this technology is of no value,'' he says. ''Sometimes we were accessing info from other units. We weren't able to get accurate readings.''

Although Miller has decided to table the project until 2006, he says some valuable lessons have been learned that could help other IT executives.

''Focus on getting information in your ERP systems up to date,'' he advises. ''It's critical if you're going to do an accurate inventory with the RFID tags.''

He also recommends creating a virtual team for the evaluation and deployment of RFID technology. ''Put together your applications, systems, production, and project management teams. Wherever you're storing the data, whatever databases you're integrating with, they're going to have to be part of it.''

Craig Mathias, a wireless expert and principal at Farpoint Group, a consultancy in Ashland, Mass., contends the real promise of RFID technology is still a few years out. But he agrees with Miller that there are things IT executives must do now to prepare for the technology.

''Probably the biggest issue surrounding RFID is database integration,'' he says. ''You want to trap and trigger the right activities stored in the databases.''

For instance, Mathias says, hospitals could use RFID tags on all of their equipment so they could quickly locate specialized tools in an emergency, making better use of high-cost resources. But for this application to be successful, he says much thought has to be put into the creation of the back-end databases. ''What are the priorities for the information being read?'' He adds that it's easy to gather too much information and not be able to sort through it.

Dan Mullen, president of the Association for Automatic Identification and Mobility, based in Pittsburgh, says the amounts of data that stream in from readers could pose challenges for IT groups, in terms of database processing, storage and bandwidth. All are concerns that have to be addressed before the technology is rolled out.

One way to save on resources is to determine what information can be handled on the front end and what should be passed along to back-end databases. ''In some cases, you might only need to see basic information, such as the type of product, so there's no need to pass along the more specific information, such as the plant it came from. In other cases, you might want more visibility and depth.''

Mullen says IT groups also have to determine how that data is going to be used and who it needs to get to. For instance, does it need to be sent in real-time over mobile devices to various departments. ''Hopefully, companies are going to use the information [being delivered], he says. ''It doesn't do any good without business intelligence.''

A final piece of advice is to know the environment the RFID tags will be operating in. Will there be interference from various materials?

Chris Heim, president of HighJump Software in Minneapolis, says IT managers must consider the engineering involved in RFID deployments. ''You can't just put an RFID tag on any substance and expect to have a high-read rate.''

He warns that liquid and metals are difficult for RFID sensors. ''If you're going to put a tag on a can of pop, you're going to have to experiment.''

''There is a tremendous amount of information that will come from RFID, but there's more witchcraft necessary than some people are used to,'' he says.

For Bill Miller, with a little time and preparation, the promise is still there: ''This is a technology that's going to be very successful in the future.''

0 Comments (click to add your comment)
Comment and Contribute


(Maximum characters: 1200). You have characters left.