Programmer, Analyst, Learner, Leader, Gamer

I am pretty well versed in RFID technology. At PREM (my last employer), we were thinking about implementing RFID for the purpose of building equipment tracking and maintenance tracking. We never implemented it due to cost with little time saved based on how we wanted to implement it, but I learned a lot about how RFID works and the value vs cost of them. So when I was asked to join a task team to identify solutions for RFID for a client at my current employer, I found this comprehensive article detailing the analytical value and ROI of RFID technology: http://www.teradata.com/t/pdf.aspx?a=83673&b=140892.

RFID solutions are not simple. They are VERY hard to implement, with a high start up cost and high upkeep cost (RFID tags 5-10 cents per tag, Barcode label less than 1 cent per label). From a physical and process standpoint they can be an operational challenge, but also from a software standpoint they can be equally complicated to implement well. The value in an RFID implementation is that they are fast and accurate in large volumes and/or frequency of scanning. The best RFID solutions are automated and don’t even require human intervention for scanning because they leverage the longer range RFID tags. Even when using short range tags which require human intervention for scanning, imagine walking into a room with 5,000 RFID tags and scanning them all in about 30 seconds (166 scans a second, some scanners are as fast as 400 scans a second). Let’s say there were 300 chemicals in the room, that’s still 2 seconds to scan all of them. Then you move on to the next room.

While they are expensive to implement and upkeep, the cost saving is in the time. From a software standpoint, you need a solution that will either POST-process all scans or buffer the scans and handle them one at a time. Ideally again since RFID is all about speed, you would want to do a lot of scanning all at once, then reconcile the scans later handling any special cases and errors all at once instead of piecemeal between scans or even scan batches. If you are slowed down by the interface of your scanning solution then it outweighs the benefits of RFID implementation in the first place.

The best RFID solution I can see is one that scans and collects data, then post scanning interfaces with a machine, processes the data, then intelligently based on pre-setup rules to handle exceptions. This is assuming that you don’t want to on the fly use their RFID tags to store information in the tag like current quantity, current location, etc, information that would be valuable to the scanning process. Who knows maybe as containers move within their facility, maybe they intend on updating the RFID tag to store actual transaction information. Of course I’m just going out on the limb of all the things you can do with RFID.

In the end, the value of RFID is all about speed. Speed, speed, and ummm speed. As long as the solution is fast and efficient, it outweighs the upfront cost of implementation and the maintenance cost of the tags / encoding devices.

Trackback URI | Comments RSS