The Truth About RFID ... as the doctor Sees It
I recently encountered yet another RFID article that took a look at how RFID will impact supply chain optimization and control and decided that I just can't ignore the subject any longer. As you have probably guessed from that first sentence, I'm not a big RFID fan. It's yet another technology that greatly over-promised and greatly under-delivered, and it did this for two reasons. ( 1 ) The promises were outlandish with respect to what the technology actually does and ( 2 ) the technology, by nature, is not as "plug and play" as the vocal proponents would have you believe.
RFID, which is short for Radio Frequency IDentification , is a method of "automatic identification" that relies on storing and remotely retrieving data using RFID tags or transponders. These tags generally contain an integrated circuit for radio-frequency signal processing and an antenna for signal receipt and transmission, however some tags are read-only and do not contain an integrated circuit for signal processing.
RFID proponents promote the use of RFID tags for supply chain use because they improve the quality of material location/movement data versus current data collection technology and this provides more accurate data delivered to existing ERP systems that drive supply chain optimization systems. As a result, forecasting, master production scheduling, and distribution resource planning can produce better, timelier and more granular outputs based on more accurate, near real-time inventory and/or material movement. In addition, RFID allows manufacturers to keep in contact with, or at least "hear" from, their material as it moves through the supply chain.
The idea is that since the tags can automatically be read by readers at each checkpoint in the supply chain, the data is immediately available for processing and immediate transmission to your systems. And even though real-time data is valuable, and the benefits of this real time data that RFID proponents promote are significant, the fact of the matter is this: you don't need RFID to know where your goods are in a supply chain and when they hit a checkpoint. You can just as easily slap a plain-old fashion barcode on every box and get the same results. Of course, you'd need someone at each step of the supply chain to scan every box, but as long as the systems are properly configured, the data could still flow automatically up and down the chain. The advantages of RFID are efficiency and human-error reduction, and that's it. It's more efficient because, as long as the scanners are properly configured, and the goods properly passed through the scan points, the tags are automatically read for the whole pallet simultaneously. It's less prone to human error, because, as long as the tag is properly attached to each box on the pallet, it will be read automatically at each point and you don't have to worry about a human missing a barcode or two each time the pallet is supposed to be scanned in full.
Therefore, RFID is valuable if, and only if the savings that result from increased efficiencies and reduced human error outweigh the costs of its implementation, which not only include the up-front equipment and installation costs, but continued costs associated with maintenance and each and every RFID chip. But it's not going to save you a fortune ... its the systems that you buy to take advantage of that extra data and optimize your chain that are going to generate significant savings, and, as a I pointed out above, they don't care if the data comes from an RFID chip, a scanned barcode, or, if you want to go back to basics, a human operator typing in a box or pallet number into a terminal that automatically propagates the data up and down the chain.
I guess what I'm saying is, before jumping on the RFID bandwagon, see it for what it is, talk to your industry peers about the increased efficiencies they saw, and do the TCO calculations before committing to it. For most operations, I doubt it's the best thing you can do to save money.
For the most part, I would agree with you (see my profile on LinkedIn, a cofounder of OAT Systems) That said - in certain applications it definitely adds value. The top two areas in my opinion are Security and Retail Softgoods. These areas, albeit combined with tag reuse does return significant ROI. Two examples - www.tjs.ae and Marks and Spencer deployment.
See this link about the failure of the DoD RFID program:
http://logistics.about.com/b/2008/10/12/new-disappointment-for-12-million-rfid-project.htm
If RFID is such a failure why have the road transit authorities used it so successfully to increase efficiency and reduce congestion for toll collection.
Any technology needs to be properly used for it's intended scope and capabilities. One shouldn't use a Porsche to help a friend move, just like an articulated dump trucks shouldn't be driven on the autobahn.
DAC:
1) I never said RFID was a total failure, just that, in my view, it is a failure in supply chain, which is the focus of this blog. As you said "any technology needs to be properly used for its intended scope and capabilities", and I do not see its capabilities as well suited to many of the applications supply chain RFID enthusiasts are trying to adapt it for. Maybe in another five or ten years, but not now. It does have some uses, outside of supply chain, and properly implemented they can be a success, as long as "success" is appropriately defined and the technology appropriately used.
2) As a trained mathematician, I have to point out that even though many transit authorities have successfully deployed it to increase their efficiency and decrease their costs with respect to toll booth operation, it doesn't always reduce congestion, and, sometimes, just like extra roads, leads to extra congestion.
While the result, in general, is counter-intuitive, I can give you a representative example if you're wiling to overlook the fact that part of the problem with this example is improper deployment.
A common initial application of this technology by transit authorities in North America is replacing most of the traditional toll booths on bridges and controlled access highways with RFID readers and RFID only lanes. Since there are usually, on average, twice as many booths as there are lanes, this sounds like a great application as first as you can typically reduce the number of collection booths to one, or two in peak periods - since not all cars will have the technology, which is a 66%+ manpower reduction.
But you need to remember that at these implementation points, the number of lanes temporarily doubles and then quickly decreases back to the original number, or less, and does so in a small amount of space, that is usually uncontrolled. RFID technology allows a large number of vehicles to initially cross the toll-are in a short amount of time, but during heavy traffic, more vehicles reach the point where the number of lanes collapses at one time than can merge onto the reduce number of lanes, causing traffic to come to a complete halt.
So, instead of having, say, 3 slow moving lanes, like you used to have before the RFID introduction (since even though the lanes doubled at the toll point before RFID introduction, the manual tolls reduced throughput to that which could continue uninterrupted once the tollbooths were passed), you have 3 lanes which move quickly, then come to a complete halt shortly after the toolbooths, and then move really slowly as human drivers, without markings or control, try to merge 6 cars into 3 lanes. And the net effect is that this fast-stop-start traffic pattern moves slower than the old pattern which allowed vehicles to essentially continually move at a slower pace.
Thus, RFID is only effective if it is deployed in a manner that does not create a choke-point in high traffic.