NPCT and the MIT Auto-ID Center
Collectively, I think we have vacillated between over- and under-estimating the importance of the MIT Auto-ID Center. IMO, it ain't Camelot, but it also ain't insignificant.
I think I have recently begun to understand both why it is potentially so important, and at the same time why the future of Nanopierce does not depend upon it.
There are two FUNDAMENTALLY DIFFERENT approaches to the world of smart tags, and there is plenty of room for both of them. I don't think anyone has really given names to these two different models of how to proceed with the development of the smart tag industry, so for lack of better terms, I'll call them the 'Real World Model' and the 'Ideal Model.' I do not intend to imply that either of these approaches is inherently superior to the other. They are merely different.
The 'Real World Model' could just as well be called the 'Tag It If It's Worth It' model. The idea here is that we begin by accepting the current state of the art in smart tag/label manufacturing, along with progressive refinements of that art. If smart labels cost 50 cents, we will stick them on things that justify a fifty-cent label. You wouldn't hesitate to put that kind of label on a $500 dress, if you thought it would reduce theft. If smart labels cost 20 cents, they would probably save FedEx and UPS a lot of money. At a dime, maybe you'd put them on a $5 package of razor blades. The cheaper the label, the bigger the market.
Obviously, there are a lot of items that are not valuable enough to justify the cost of a five-cent label. But even at 20 cents or so, there is a market for billions of smart labels per year. Philips, TI, Infineon and others are counting on tremendous growth in the demand for their RFID chips as the cost of smart labels is steadily driven downward. I believe that Nanopierce's NCS will play a significant role in this process by helping to cut costs and improve quality.
While twenty-cent and ten-cent and even five-cent labels have many extremely important uses (tracking and inventory, package delivery, airline baggage, prevention of theft, etc.), this 'Real World' approach has a very important limitation in comparison with Bar Code labels: As long as there are a lot of items that are too cheap to justify the cost of a smart tag, these labels will not be adopted for use at retail cash registers. That's why there is room for a second approach to the question of smart labels.
The 'Ideal Model' might better be called the 'Tag EVERYTHING Model,' and it is this approach that is firmly entrenched in the MIT Auto-ID Project. Why? Several reasons, I think: First, there is a demand for it. Supermarkets, drug stores, retail stores of every variety would benefit from being able to ring up merchandise quickly and easily using RFID labels. Second, MIT people like to think big. They are big-time 'systems thinkers,' and the bigger the system, the better they like to think about it. So they envision a world in which EVERY object manufactured has an RFID tag attached, which not only would allow for universal use in checkout lanes, but which also goes way beyond that, to things like the appliances of the future where your refrigerator 'tells' you that those eggs are past their expiration date and your microwave 'reads' information from your frozen food and cooks it accordingly. Third, take a look at the end-user sponsors of the Auto-ID Project. Gillette, Proctor and Gamble, etc. etc. They get a lot of sponsorship money from big-time consumer products makers who would like very much to be able to put smart labels on everything they make and ship and sell.
When you begin with the goal of putting smart labels on EVERYTHING, then the cost of the labels is all-important. In order to do this, you have GOT to be able to get the cost down to a penny (or two at most) per label. That's gonna be pretty tricky, given that the real estate on a Silicon wafer costs about four cents a square millimeter. So this in itself dictates that your chips can't be bigger than a quarter of a square millimeter. Chips that small are quite difficult to handle using standard 'pick and place' methods, which explains the Auto-ID Center's love affair with Alien Technologies. (Alien has a process called 'Fluidic Self-Assembly' which enables them to place monstrous numbers of extremely small chips on a substrate very cheaply.)
In addition to the cost of the silicon itself, the complexity of the circuitry on the chip is an important cost factor. The more logic and memory you put on the chip, the more it costs. So, the Auto-ID Center philosophy is to strip the chip itself down to the absolute barest minimum. All they want to have on the chip is 96 bits of memory which can be read as many times as you want, but written to only once. Their version of an RFID chip, to put it bluntly, is not very smart. All the chip has is a big, long, unique identifying number. The real intelligence is in the network of computers off the chip. When the chip is read, the reader has to be able to access a database in order to find out that item number such-and-such is a 16-ounce container of BBQ flavored Pringle's priced at however much the store is charging. Presumably, your local Kroger's would receive data from all their suppliers telling them what merchandise corresponds to what range of item numbers, and they would maintain their own prices.
Actually, the 'Real World' players have been going in the completely opposite direction: They have been making their chips smarter and smarter. Gigantic amounts of memory which can be rewritten as many times as you please, with encryption capability and partitioned into segments so that only the manufacturer can write to this segment, the shipper to this segment, the retailer to this segment. Bigger, smarter tags have applications that the el cheapo versions envisioned by MIT can't even dream of.
Thus we have two FUNDAMENTALLY DIFFERENT ways of approaching the implementation of smart label technology, each driven by somewhat different goals and philosophies. IMO, it is very clear that NCS has an important role to play in advancing the 'Real World Model.' I think it is a lot less clear as to whether NCS will play any role in the 'Ideal Model,' since this depends on things that are not known at this time. In particular, we do not know whether Alien Technologies has developed a method of attaching antennas to their fluid-assembled chips without needing NCS, or, OTOH, whether Nanopierce might offer them the best method of solving this problem.
In any case, I don't see how we could possibly view the MIT Auto-ID Center as insignificant, given the enormous size of the consumer products companies that are involved as sponsors. (How big are they? Here's one way to look at it: If all the sponsors of the Auto-ID Center were to use smart tags on all their products, they would consume roughly one tenth of the world's supply of silicon wafers, possibly a bit more -- and that's using very optimistic assumptions about how many chips can be made from each wafer.) The opportunity to rub elbows with companies like International Paper, Gillette, and Proctor & Gamble, with Noel Eberhardt making the introductions, is priceless.
OTOH, there may be a very legitimate possibility that the visions of the MIT Auto-ID Center may be little more than pie in the sky for years to come. They have a pilot project in place right now to evaluate the use of smart tags in the retail supply chain. Very big players (manufacturers and retailers) are participating, and I imagine they are probably actually putting smart labels on every little package of gum and running them through RFID readers at the checkout counters. But I flat-out guarantee you that those labels were not made for a penny apiece. And until they can be, I do not think that RFID labels will attain universal use in the retail supply chain.
In the meantime, the 'Real World Approach' will continue to make steady progress. Billions of labels will be sold for 15 cents, and billions more for a dime. People will get filthy stinkin rich, and IMO NPCT shareholders will be among them. So, maybe we'll become an integral part of MIT's vision of the future, or maybe not, but I think NPCT can make a lot of money with or without being adopted by the MIT-Alien crowd.
Best regards,
Geoff
