By Mark Roberti
June 15, 2009—I spoke at the recent U-Connect conference hosted by
GS1, offering the audience an overview of
radio frequency identification adoption around the world, as well as in different industries. After my presentation, a gentleman from a well-known retailer came up to me and rather sternly asked, "Can you guarantee that I will be able to
read every
tag, every time?" I told him that I could not, nor could anyone else. "Then this technology is useless in retail," he said. "How can we manage our business if we are always not accounting for some of our inventory?"
For a moment, I reacted with a stunned look. After recovering, I replied, "But that's what you're doing now." He didn't understand, so I pointed out that he's currently not accounting for every item in his store's inventory. In fact, the inventory counts in his store are probably only about 65 percent accurate. But he continued to insist that
RFID needed to be perfect before it could be deployed in a retail setting.
The only time RFID needs to be perfect is when a financial transaction is involved. And in those cases, it is perfect. A few years ago, I visited
TransCore's test facility in Albuquerque, N.Mex. The company has a test track where cars and vans of difference sizes zip around a track to test the reliability of its toll-collection systems. As far as I know, these systems work flawlessly.
Exxon Mobil's Speedpass and
MasterCard's PayPass work perfectly as well.
Reading tags on products or assets, in most cases, is a different matter. RFID doesn't need to be perfect. It simply has to deliver a
return on investment (ROI) that is compelling enough to warrant adoption. Does it do that?
American Apparel,
Charles Vögele and other retailers have found it does (see
Charles Vögele Group Finds RFID Helps It Stay Competitive and
American Apparel Makes a Bold Fashion Statement With RFID)—and the reason it does is that retailers have great difficulty keeping track of their stock.
Bill Hardgrave, director of the
University of Arkansas' RFID Research Center, will present aggregated data at our
RFID in Fashion 2009 event (being held in New York on Aug. 12-13), illustrating that overall in-store inventory accuracy is off by 35 percent. RFID can help bring that level up to 98 percent or 99 percent—and at the conference, we'll demonstrate how. Yet, the retailer at U-Connect was more worried about the few items RFID can't read, rather than the fact that his inventory accuracy is appallingly low today and would be much better with radio
frequency identification.
READERS' COMMENTS
People, Process, Technology
Mark, You make some excellent points in your Editorial Note. However, you could have expanded your argument to the more general case to state that no data capture system is "perfect", and that your argument is not confined to retail operations; yet the nature and precision of data relating to inventory (count, locations, additional meta-data) continues to increase, and the ease (degree of automation) and accuracy associated with such data capture exercises has improved dramatically this century as a result of multi-faceted technology innovations. Lastly, never forger that there is always an essential need to strike the right balance in the people, process, technology mix that constitutes the implementation of any AIDC solution. What we observe is an increasing reliance on the technology component of this equation, but it would be foolish to ignore the requirement for rigorous processes, both as a matter of routine operations and of course exceptions handling. Best regards, Chris
Posted By: C. Hook 6/19/2009 at 7:15:46 AM