Here's an older but interesting article.
It gives a decent general overview from April 2002 but
its useful because it helps understand timing for RFID.
There are different cost structures for different players.
That is while you'll see the article mention 7-5 cent chips
as being an early adaptor point and <5 cents as generally
cost effective for the majority - the reality is that
there is a range of cost effectiveness depending on the
industry and to a large degree the relationship between
the full cost of the item to be tagged and cost per tag.
(RFIDing packages of bubble gum playing cards will take
longer than RFIDing packages of Synchron shavers).
[edit - also the full incremental cost of implementation
spread over the number of uses]
The Gillette/Alien news is the bleeding edge of the broad
adaptation of RFID which the next 48 months will show.
They will be begin shipping the tags in March.
If you look at their site and the articles referred to you
will see that Alien is now producing the first sub 10 cent
chip.
alientechnology.com
That's why Gillette which was one of the originators of
the AutoID Centre is going ahead (with selected product
that fits the cost/benefit profile).
The comments about Gillette's strategy above are my opinion.
Here's the article:
New chips could make everyday items 'talk'
By Kevin Maney, USA TODAY (04/11/2002 - Updated 10:34 PM ET)
Home uses for ePCs
Kitchen: A frozen dinner could transmit cooking instructions to a reader inside the microwave.
Laundry: A sweater could tell the washing machine its fabric care instructions.
Medicine cabinet: The drugs could talk to each other and post alerts if they find a combination that could cause a dangerous interaction.
Home office: Filing cabinets could know which tagged documents are in which drawer, transmitting the location to a computer screen.
Entertainment: A music buff could browse a list of his CD collection and see which ones were in the house or the car.
In Singapore, cars "talk" to the streets they drive on.
In Tulsa, retailers test a system that lets products inform the store when they're bought.
In home kitchens later this decade, frozen dinners might automatically give cooking instructions to microwaves.
The Internet revolution was about people connecting with people. The next revolution will be about things connecting with things.
And it's taking shape in pockets around the globe. For the first time, big players such as Wal-Mart, Gillette and Procter & Gamble are joining to give the technology serious momentum.
In a twist, this next technological chapter won't emerge out of ever-more-powerful computers and faster Internet connections. This shift comes from the opposite direction.
It will ride on pieces of plastic the size of postage stamps, costing a nickel or less. Each tag will contain a computer chip, storing a small amount of data, and a minuscule antenna that lets the chip communicate with a network.
In time, when billions of tags are out there and communicating, the technology will infiltrate business and everyday life to a greater extent than today's personal computers, cell phones or e-mail. In decades to come, its impact might be as fundamental as the invention of the light bulb.
Those tags will someday be on everything — egg cartons, eyeglasses, books, toys, trucks, money and so on. All those items will be able to wirelessly connect to networks or the Internet, sending information to computers, home appliances or other electronic devices.
Grocery items will tell the store what needs to be restocked and which items are past their expiration dates. The groceries will check themselves out in a split second as you push a full cart past a reader. A wine lover could look on a computer screen and see what's in her wine cellar. Prescription drug bottles could work together to send you a warning if the combination of pills you're about to swallow would be toxic.
"Any single one of these (tags) is like a one-celled organism," says Glover Ferguson, chief scientist at consulting firm Accenture. "They're just smart enough to say their own name." Like cells, their power will come from billions of them working together, he says.
"We're really talking about the next 50 years of computing," says Kevin Ashton, executive director of the Auto-ID Center at the Massachusetts Institute of Technology.
Auto-ID is the program backed by Wal-Mart and the other blue-chip companies, and the center is trying to create a standard, like Internet protocol, for the tags' communication.
That would enable any tag to connect to any network, much as any PC can work on any network.
The technology doesn't really have a handy name. The tags are known as radio frequency identification tags, or RFID. The Auto-ID center calls the core of its standard "ePC," which stands for Electronic Product Code.
Perhaps an appropriate umbrella name might be tinyband. Today's hefty computers and super-fast fiber-optic networks communicate on broadband technology. Tomorrow's little nickel tags will work on tinyband technology and as little as 96 characters of information.
Reducing cost is key
RFID has been around awhile. During World War II, the military used a high-powered, bulky version of it to identify friendly aircraft. Starting in the 1970s, the federal government stuck RFID tags on nuclear materials to better track them. In the 1980s, commercial warehouses used it to locate loaded pallets.
These days RFID shows up in a few familiar places. The technology is in ExxonMobil's Speedpass — a key fob that works like a credit card, wirelessly identifying you to a gas pump. On highways across the USA, wireless toll booth systems such as E-ZPass work on RFID.
Singapore relies on the technology to control traffic. Its system, called Electronic Road Pricing, or ERP, charges different prices to drive on different roads at different times. Driving on one main artery between 8:30 a.m. and 9 a.m. costs $3 (in Singapore dollars — about $1.60 in U.S. currency) but is free from 2 p.m. to 5:30 p.m. The pricing encourages drivers to stay off busy roads at busy times.
Every car must have an RFID tag, which communicates with readers along every major road. The road readers identify each car and send information to a central computer, which adds up car owners' bills.
Until now, the tags have been too expensive for anything but specialty applications like E-ZPass and Singapore's ERP. One tag costs about $1 — hardly worth pasting to a $3 frozen dinner or even a $20 bottle of wine.
But a small, private California company called Alien Technology is pioneering mass-production methods that will radically reduce the cost. Later this year, Alien will take orders for 500 million tags at a time, selling each tag for just under 7 cents, says Alien CEO Jeffrey Jacobsen, who notes that one such 500 million-tag order would exceed all the RFID tags ever made.
The Auto-ID center figures the tags must get down to 5 cents each. "That's kind of the price point at which CEOs don't automatically throw you out the door," Ashton says. Jacobsen says tag prices won't drop to 5 cents until at least 2005.
At 7 cents, major companies consider the technology promising. At 5 cents, it would start rolling out into business applications.
If manufacturers and retailers can get accurate, instantaneous information about where products are and how many are sold or delivered, they could streamline operations and save millions of dollars a year.
"This idea is seeming less and less crazy and more and more desirable," says Auto-ID's Ashton. "Technological breakthroughs show it's not only desirable, but inevitable."
Widespread consumer use of tinyband will take time — perhaps a decade or more.
That's what happens with new technology. Computers didn't move from businesses to homes until more than 30 years after the technology was born.
Don't expect it overnight
Some chastise tinyband proponents for promising too much too soon. "You have to manage realistic expectations," says Cliff Horwitz, CEO of SAMSys, which is making a universal reader that can talk to tags from any manufacturer. For the foreseeable future, "MIT has a pretty extreme and unrealistic view of the world."
Others, though, can't contain their excitement. The real fun will start once the price of a tag gets down to around a penny. Then adding a tag would be no more expensive than stamping a bar code on a product. Bar codes today are on nearly every item made for consumers and business.
Imagine that every one of those things will have a small amount of intelligence and ability to communicate. The world around us would almost come alive.
Lost glasses? Tag to the rescue
Arno Penzias — a Nobel prize-winning scientist, one-time head of Bell Labs and an investor in Alien Technology — has a favorite microcosmic scenario:
You lose your eyeglasses. They've fallen under the family room couch.
The tag on the eyeglasses connects with a reader in the family room — readers would be all around a house. The reader is also getting signals from everything else in the room.
Tags work a little like radar. A reader sends out a signal looking for tags. The signal excites the tag — the tag itself has no power — and causes it to return a signal containing its information. This request and return of a signal happens more than 100 times a second for each tag.
The reader pipes its information across a wireless network and dumps it into the home computer. The computer looks at the data and deduces that the signal from the glasses takes the same amount of time to hit the reader as the signal from the couch.
You sit at the computer and type in a search box: "Where are my eyeglasses?" The computer spits back: "Under the couch."
"In a few years, high-end consumers will likely start using tag readers to locate items in the house," Penzias says.
Into commercial uses
On a more practical level, the industry is watching a tinyband test in Tulsa.
Several stores and manufacturers agreed to put tags throughout the supply chain, so the tags are on crates of products, in trucks, at loading docks and all around warehouses. The companies involved are being kept secret, though two seem to be Wal-Mart and Pepsi-Cola. Industry watchers say the technology has been working better than expected.
The Auto-ID Center figures that testing will go on until 2003, when the technology will start to flow into commercial uses. In 2005, it will start to be widely adopted by business. Early-adopter consumers could bring tinyband into homes around that time.
A study by research firm Venture Development found that RFID will be a $1.4 billion industry in 2002, climbing to $2.6 billion in 2005. If so, it will still be only a speck in the overall technology sector.
Aside from technology challenges, tinyband will increasingly test society's acceptance. Privacy will certainly be an issue.
For instance, insurance companies might want to use the technology to know where you take your car, so they can charge more if you regularly park in high-crime neighborhoods.
Privacy "is an issue. There will have to be a social discourse about what we want and don't want, " says Accenture's Ferguson. "But the technology isn't going away: You can't un-invent it." |
