RFID applications in business (as distinct from military, security, etc,) can be roughly divided into two categories, business or consumer related. Business related applications are mainly supply chain oriented, although other logistic and distribution systems could benefit. Consumer related applications are all about detailed marking of individual goods items.
There is a huge difference in the volume of tags needed for the two categories. Supply chain management (SCM) may include individual items but in most cases it will be containers, palettes and cases that must be tracked which is a relatively small volume of units compared to the items stacked on shelves. Also the unit value is very different, which has a big impact on the cost effectiveness of RFID. There is however another major difference and that is the social implications. SCM is a B2B system and as such does not impact directly on the consumer, but individual items purchased by individual customers do and that is another matter. Already there are consumer protest groups raising strong objections to the use of RFID tags. The source of the problem is not in the store, but outside it. The argument is that any goods with an embedded RFID tag can be detected anywhere, considered an invasion of privacy. This of course is nonsense at the moment because the tags are too primitive to contain any details of significance and the goods have to be held close to the reader. But imagine the future when features similar to active tags are available with low cost passive tags and the details of the customer can be detected from a credit or loyalty card and added to the tags on the individual items - and hence outside the store. One proposal is to implement an RFID reader system which can destroy the tag contents on each item as it passes beyond the checkout, but this is no mean task. A better concept being explored is to selectively erase data in the tag according to various criteria, rather than disabling the whole tag.
The current cost of tags will to a large extent govern the rate of acceptance. Inevitably tagging of high value items will be more acceptable and hence the SCM systems will take off first, for which we must be grateful! Simple tags (higher cost active tags are more specialised products) today cost around 5 to 10 cents. It is predicted that they should be in common use by 2008, with a unit cost down to 1 t0 2 cents. Personally I am disappointed that they are not in common use on pallets and cases already. It is pointless predicting further than that.
One other technical development of interest is the combination of tags with other sensors, in particular temperature. The shelf life of perishable food stuff and other goods is often affected by the ambient temperature in which they are stored. By detecting and recording temperature samples a longer useable life will be possible (not to mention less chance of food poisoning!). There will be other similar examples, some of which would cost justify more expensive active tags.
It probably won't be a major problem but the issue of who pays for the tags and readers should be addressed early. Don't forget all the aggravation that accompanied the introduction of point-of-sale funds transfer; the banks assumed that the retailer would pay and the retailer assumed that the banks would. Sort it out before it becomes a problem.
In conclusion, "keep it simple, stupid" to begin with; concentrate on reducing inventory with SCM before product tagging.< BR>
Martin Healey, pioneer development Intel-based computers en c/s-architecture. Director of a number of IT specialist companies and an Emeritus Professor of the University of Wales.