Ready For RFID Gen3?

Few are, but as item-level tagging looms on the horizon, security and functionality may stretch Gen2 beyond its limits.

Considering the glacial pace that is the world of global standards bodies, it's not surprising that developers need to think far ahead. But as the politically-complicated RFID community has just started getting used to RFID's Gen2, some are already prepping for Gen3. That's making some in the industry feelings a little skittish--or perhaps nauseous.

"Some people are so committed to RFID Gen2" that talk of Gen3 "is like garlic to a vampire," said Kevin Ashton, a VP at ThingMagic and co-founder of MIT's Auto-ID Center.

Ashton argues that if past RFID standards effort are used as a benchmark, it will take between two to four years to come together on a Gen3 RFID standard, which would put it into the 2008-2010 timeframe, which is not coincidentally "when I think item-level tagging will get real volume."

Volume item-level tagging and the processes that will support are going to put huge demands on RFID systems and Ashton doubts today's chips will hold up.

"What people are talking about is security and item level tagging. There is some dancing around the inevitable conclusion that we need an improved protocol to address this," he said. "Some of the dancing takes the form of attempts to shoehorn security into Generation 2 or to deny that there is a problem. But if you don't want eavesdropping, and you want to authenticate tags and readers, you need so many changes to Generation 2 that it may as well be called Generation 3. There are some other security features that would be nice to have too - like non-repudiation: undeniable evidence that a certain reader read a certain tag."

None of this should be a surprise, of course. "Lots of people get that there is going to be Generation 3 and we're getting some kudos for saying so publicly," Ashton said. "At the same time, there are people who would love Generation 2 to live forever. Same thing happened with Generation 1. The more things change, the more they stay the same."

This gets into the never-ending semantic debate about when does software morph from a patch to an upgrade to a dot zero release. How many changes can be done to Gen2 before it truly is Gen3? The good news is that there are plenty of standards meetings to debate such matters. The better news?for all concerned--is that I won't be attending any.