Not Just a Few Bytes Anymore
Fri, 03/05/2004 - 10:27am
Ern Worthman, Technology and Editorial Director
Not too long ago and in another life I wrote a column about something called "auto-dining." The participants in the program were consumers and restaurants. The consumer would carry a RFID "smart card" that would automatically transmit data to the sensor at a participating restaurant. On the smart card was information about the consumer's dining habits, likes and dislike. For example, the card might contain the patron's favorite cocktail (shaken, not stirred if you please) or the number of lumps of sugar in their favorite Tarbucks concoction (sign me up!).
Well, I smelled opportunity then and, now, I smell it again. Recently, I came across an article in a magazine (www.pcworld.com the consumer watch section of the magazine) that tells me RFID (like 802.xx ) is a technology to watch.
Why? Well, first of all, consumer computer mags are picking it. Second, it is becoming less and less significant as a BOM line item. (target: five-cents per tag at market maturity). Next, it's getting really, really tiny (some chips are the size of a grain of salt!). These perturbations, and others, look like a window of opportunity to me.
For the longest time, RFID was a low bit count, short-range technology relegated mainly to inventory control. Well, not anymore 2004 is set to be the year RFID rises from the ranks of the proverbial ugly stepchild.
Witness the fact that Wal-Mart and the DOD have set the mandate that all of its major vendors must be RFID-literate by January of next year. Metro Group, the world's fifth largest retailer, put out the word that ALL of its supply chain vendors must be RFID capable soon as well.
Hearing the call, giants like IBM, Microsoft, SAP, Sun Microsystems and Intel are all jumping on the bandwagon, as are Texas Instruments, Symbol Technologies, NCR, and Philips. Some of these players are software, some are hardware, and all are serious, big-name players.
One project teams IBM and Philips Electronics. The companies plan to offer RFID solutions not only for supply chain management, retail and asset management, but smart card solutions for finance, e-government, transportation, and event ticketing. Eventually, they will create products and services based on their new system and use that technology to target new customers. Philips is aggressively pushing into the wireless market, testing new products and developing new services such as a wireless, contactless payment system with Visa. Even beef is a candidate, via a plan to tag each cow with a RFID tags (oughta be pretty easy to follow that mad cow, right?).
One example of a technological innovation pushing RFID into new markets is placing the coil (the biggest roadblock to ultra-miniaturization) on the chip itself. This makes the chips more robust against deterioration of contacts and mechanical stress, because there is no external connection between antenna coils and chips. This has vertical and horizontal markets written all over it (those that have a bit better ROI than asset/warehouse management).
Furthermore, technical advances with these products include 4KB and up of memory and multiple levels of security. And they can now be read at over 100 MPH (I'm gonna test that at the next turnpike toll booth). Another significant advancement is that these R/W LSI tags are not just read-only anymore. They are writable (so now I can also automatically change my Tarbucks coffee preference).
All of this sounds exciting, for a relatively dull market segment. But, just as there is no free lunch, cheap and robust RFID technology isn't just waiting around to be scooped up. There are some roadblocks standards and privacy issues for starters. And, that five-cent tag will only be a reality when they produce billions monthly, and then only for the very basic tag. Most usable tags are still at a half-buck or so with specialty tags higher.
On the standards issue, the players seem to be lining up one of two sides. The DOD likes ISO, while the commercial side wants the standard being devised by a company called EPCglobal Inc ( joint venture between Uniform Code Council Inc. in Lawrenceville, N.J., and EAN International in Brussels. There is also the EAN.UCC GTAG initiative floating around out there). That could get ugly (read expensive) if something doesn't give since many companies supply both governmental and private sectors.
The security issues have some basis in potential, but the paranoia that RFID tags on merchandise will violate the consumer's privacy by "spying" on the individual and collect data is unfounded. Rumors that these tags can be tracked by satellite are ludicrous since the range of these tags is only a couple of meters, and they do not contain any personally identifiable information. It's far more productive to hack a credit card database or snoop on wireless conversations. Besides the infrastructure to track tags with only a few meters of range is unworkable (unless we've got some secret technology with receiver sensitivities 200 dBm).
Finally, while the tags may be getting cheaper, the above-mentioned infrastructure isn't. The cost of installing smart warehousing and ubiquitous tracking on a world wide scale is still astronomical, so economies of scale play large here.
Never-the-less, the mandates are there. One has to assume that some of these mandates will slip, but make no mistake, they will happen. Depending upon the factors I've touched on (and some others I didn't have room for), and the usual unforeseen disasters, conservative estimates put the tag market alone at over a billion dollars by 2008 the optimist look for it to be three billion, plus.
So, I guess if I were the CEO playing in the wireless arena, looking at that five-year plan, at a minimum, I'd keep it on the radar screen and be ready to move.