Do You Need 100 Percent Read Rates?

By Mark Roberti

For some applications the answer is absolutely 'yes,' while in others, achieving 95 percent or higher is more than sufficient.

After I posted my last column, "How to Burnish RFID's Image," a friend in the RFID industry reached out with some comments. One was that users expect 100 percent read rates all the time—it's a digital technology and it should work perfectly. My question is whether this is a perception that the industry needs to address, or a reality.

When I speak to retailers, I often show videos of people taking inventory counts of hundreds of items within less than a minute. I am then asked, "How do you know if you have read every tag?" I usually respond with a question of my own: "How do you know someone reading barcodes didn't miss any items?" The truth is that you don't. People miss more items with barcodes than when taking inventory via RFID. But the perception is that you'd be more accurate picking up each barcode and scanning it than you would be waving an RFID reader. Studies show RFID is not only far quicker, but far more accurate.

Now, you might miss a tag that is blocked by another tag, but in a retail setting, reading 95 percent of the tags is sufficient. Why? Because taking inventory counts weekly and achieving 95 percent of the tag reads allows you to reduce out-of-stocks and increase sales. Spending more money on labor to take inventory daily can get your inventory accuracy up to 99 percent, but the increase in sales might not be offset by the increased labor costs—and by the way, your inventory accuracy currently averages about 65 percent with barcodes, so RFID is a massive improvement.

Some applications do require 100 percent tag reads. Let's say you are reading tags in surgical sponges to ensure none are left inside a patient. Reading less than 100 percent is clearly unacceptable, and the systems on the market are able to achieve that number. Toll collection is another obvious application for which you need to read 100 percent of tags. Missed tags are lost revenue. Again, the systems on the market are able to achieve this.

The same goes for ski resort applications, for which tags in lift tickets must be read. There are many other applications for which 100 percent read rates are absolutely necessary as well, and many for which it is not. RFID will never be perfect. You will never read a passive tag on a can at the center of a pallet full of cans.

The industry needs to change the perception that this means it is not a viable technology in many applications. That is simply untrue. In most cases, RFID systems can be set up to work effectively, or to achieve 100 percent accuracy if that is required. For example, you can read the tag of each can being placed on a pallet, and then associate the tag IDs of all the cans on that pallet via a single pallet tag. Thus, reading the pallet tag would give you information about every can stored on it.

I believe perceptions about what is needed from an RFID system, as well as the cost and the potential ROI involevd, are holding a lot of companies back from investing in RFID. The industry has not done an effective job of addressing these issues.

Mark Roberti is the founder and editor of RFID Journal.