There is a need for multi-frequency
interrogators for RFID
tags. This is because there will never be one ideal frequency for all applications and the protocols - the way the data is transmitted - will not be simplified rapidly by the profusion of incompatible standards and proprietary technologies that are evolving. The very versatility of RFID creates the need for universal readers and maybe even universal read-writers - the two basic types of interrogator.
Thus Wal-Mart will insist on one type of device to read most types of tag. AWID and ThingMagic of the US and SAMSys of Canada are among those developing such devices, initially for reading several types of tag at one or two frequencies only, with or without modules for versatility of customisation.
There are unique challenges in making interrogators for ultra high volume Total Asset Visiblity (TAV
), notably The Internet of Things
. This is based on widely dispersed readers in unprecedented numbers, ultimately reading trillions of tags many times, perhaps millions of times or more, every year. Even then, talk of real time monitoring all the time is wide of the mark for the biggest potential, the FMCG
supply chain and even for other supply chains. It may be possible for static archives in museums, art galleries etc. but rarely when these are transported for renovation, loan etc.
envisage hundreds of millions of interrogators being deployed. At best that gets us real time some of the time. It implies twenty million or so being made yearly at a cost, say EPCglobal, of no more than one hundred dollars each (a figure of $10 is sometimes mentioned). These figures could be realistic given that supermarkets and warehouses may need about 100,000 interoogators each, largely in the form of smart shelves
, but also at checkouts or on trolleys (so there is no checkout) on fork lifts, trucks and so on. Few would be hand held units, because the whole idea is automation.
Although a number of putative suppliers feel they can make ten dollar readers at these volumes, despite the fact that twenty million per year is actually a small volume in terms of economy of scale of electronics production, these would be primitive readers. Indeed, since no read-write tags are envisaged these need to be just read-only interrogators. However, EPCglobal
agnostic, advocating that the optimal tag should be used for various applications. For example, a large package may need a long range and therefore relatively large tag at say, UHF
, but a tiny product will need a very small tag, not necessarily possible at UHF. There is also a problem with UHF where liquids are present - a commonplace in supermarkets. All these factors present challenges to keeping reader costs down.
In fact, there are four aspects that drive up the cost of interrogators for The Internet of Things
2. Radio or hard-wired connection to the network in order to give real time at least some of the time.
3. Anticollision (reading many tags in the field at one time).
4. The unprecedented and unmanageable volume of data generated, much of which is not needed, means the interrogators must process information before passing it on in simpler form.
Something has to give.