Aquarum Vita

A Treatise On RFID

February 24, 2006 01:26 AM

Late last semester, I began research on Radio Frequency Identification (RFID) under Professor Papachristou. I wasn't really all that familiar with the technology, so I had a lot of reading to do on the subject. While I am still learning new things about it every day, I think that I'm finally to the point where I can put together some coherent thoughts about where RFID is today, and where it is going in the future. Mind you, my opinions may be flawed in some way, and you are welcome to point this out to me for the entire intarwebs to see. Ah, the wonders of technology! So, without further delay, here is my current feelings about RFID. And you probably believed that I would only write about beer!

I think that RFID is a wonderful technology. Radio frequency communication is incredibly mature, and the idea of tracking things that are out of the line of sight is incredibly desirable. Tags are going to become ridiculously cheap in the near future, and once things become a little more standardized, readers may follow suit. Wal-Mart, Target, and many other huge corporations are pouring tons of money into RFID. It may not be very long until we see tags on everything.

The ubiquity of RFID tags, however, is my nightmare. Scream "tin foil hat" all that you want, but just like any other computer system, a poorly planned implementation will wreak havoc on its environment. And the environment that we are talking about is every single item in the world (or at least those parts of it with the money to use the tags). In fact, even a carefully thought out plan will have its holes- just ask any computer security expert. I feel that RFID tags should not pervade the consumer market on a wide scale, which is exactly what industry experts believe will happen.

Think about it: why does your Snicker's bar, or your pants, for that matter, need an RFID tag? The short answer is, it doesn't. At least not as far as you are concerned. For the companies, they make a lot of sense- better tracking of inventory, the ability to tell where things are in the supply chain, or ease of identifying items slated for a recall. This, of course, could potentially lead to lower prices for consumers (although the cynical side of me says that it'll actually lead to higher profit margins for the companies). Of course, the simple solution to this is to kill the tag as it leaves the store.

But it isn't really that simple. Companies would love to cut down on scams, such as stealing an item and returning it for store credit. Having worked in retail for some time, I can vouch that this is a very real, and very common occurrance. RFID tags solve this problem by allowing the store to know whether that pair of pants has been sold or not, and the system would be completely automatic. If it isn't listed as being sold, then no return. This system requires live tags, though. It also allows companies to keep track of return customers, and also to direct advertising at them as they move through a mall (like in Minority Report). It is in the companies' best interests to keep the tags alive.

I think that this is why we keep hearing all of this talk about smart refrigerators that tell you that your milk is sour. It is to make us accept the ubiquity of RFID under the guise of some consumer benefit. I'd argue that there is no consumer application that would require the use of RFID tags on every single item that you own. It may be useful in some cases (think TV remote and keys), but the fact remains that the proposed consumer applications are extremely trivial. And this is assuming that the tags are 100% reliable in all cases, which is far from the truth. The backscattering of radio waves from the tags to the reader is by no means fool proof, and is in fact very sensitive to all sorts of environmental conditions. You'll have to pack those winter clothes very carefully to find that hat you were looking for when the time comes.

I suppose the main bullet point here would be that I like the idea of RFID technology, but not pervasive implementations. It leaves too much room for problems. There are a ton of security issues, and from the research I've done, I've come to the conclusion that they are not all that far-fetched. Personal tracking is incredibly easy with all current systems, and Denial of Service attacks are another real threat. I'll get around to writing a more detailed entry about security, but I'll save it for another night. You'll probably just have to take my word for it for now.