RFID Read Reliability Is Not an Issue

By Mark Roberti

RFID is far more accurate than existing systems, yet some people continue to insist they can't adopt the technology until it is 100 percent reliable.


I was speaking with a researcher the other day who had been running some tests on the ability to read radio frequency identification tags on a wide variety of individual consumer products in random orientation. The rates were very high—98 percent or better—but the researcher said he was afraid to publish the results because opponents would say, “See, you can’t use this technology—you’ll miss 2 percent of every item scanned.”

The researcher’s concerns are justified. A lot of retailers, logistics companies and manufacturers express concerns about RFID not being able to read every tag, every time. I’m not sure if they’re using this as an excuse not to deploy the technology, or they just don’t understand that RFID is far more accurate than either manual counting or bar-code systems.

I know some opponents of RFID just jumped out of their chairs, screaming: “RFID is more accurate than bar codes? Is Roberti nuts?”

No, I’m not nuts. The truth is, manual counting—either simply counting or scanning a bar code on each item—isn’t close to 100 percent accurate. If you think manual counting is accurate, spill a large jar of coins on the floor and ask 10 people to count them. You’ll probably get 10 different numbers.

For anyone who thinks bar codes are more accurate than RFID, I suggest you run this test: Build two pallets. Put them on pallet jacks and wheel them both through a portal. See how many RFID tags you read, and how many bar codes you read. I guarantee you’ll capture more RFID data than bar-code data.

RFID is clearly superior to current systems used in retail apparel stores. The University of Arkansas’ RFID Research Center conducted a study at several Dillard’s stores to test RFID’s ability to improve inventory accuracy. During the study, a regular inventory count was done in the store by scanning items with bar codes. When RFID was used, inventory accuracy improved by 4 percent.

I don’t mean to disparage the bar code. It’s an extremely useful and important auto-identification tool. But it frustrates me when people suggest they can’t use RFID because it can’t read every tag on every item or case. RFID technology is very reliable, but it has to be used appropriately.

No manufacturer would try to read every bar code on a pallet going through a dock door. But many people insist RFID has to be used that way. In fact, manufacturers can collect the information they need to improve their operations by reading tags on cases—hands-free—when those cases are being stacked on the pallet.

The issue is not whether RFID is perfect. The issue is whether you can use an RFID system to capture data that can be utilized to create more value than the system costs. In apparel retail, a single handheld interrogator, coupled with fixed readers at the receiving area, the replenishment area and the door between the back room and retail floor, can boost inventory accuracy from 60 percent to better than 90 percent. Some people don’t want to even consider the return on investment (ROI) that would deliver, because they are fixated on the fact that RFID might miss two or three of the 10,000 items in the store. And people think I’m nuts?

The only time RFID needs to be 100 percent reliable is when a financial transaction is involved. In most cases, it is. Visa and MasterCard trust the technology enough to use it in credit cards, and the technology has also been employed in Mobile Speedpass cards for years.

Will you be able to walk through a checkout portal without stopping, like in an IBM commercial of a few years ago, and have all of the tags read? I don’t know. Given the wide variety of items you could purchase at a mass-merchandise store, and the possibility that one metal item might shield a tag on another item, it might never be 100 percent accurate. Even here, it’s not clear that bar codes are 100 percent reliable, or that RFID needs to be. If using RFID at the point of sale increases revenue by $1 million, would you really worry that it missed $10,000 worth of items?

For companies to evaluate whether RFID will benefit their operations, they need to focus on the ROI it can deliver. RFID Journal is developing an ROI calculator for retail apparel, which will be given away at our upcoming RFID in Fashion event, to be held on Aug. 11-12, 2009. And we plan to create calculators for health-care providers and other industries as well. Our hope is that these will help companies estimate the potential ROI and run pilots focused on capturing information that can be used to assess the actual ROI, instead of spending all of their time testing whether RFID can read every tag, every time.

Mark Roberti is the founder and editor of RFID Journal. If you would like to comment on this article, click on the link below. To read more of Mark’s opinions, visit the RFID Journal Blog or click here.