CNET también está disponible en español.

Ir a español

Don't show this again

Tech Industry

RFID: Is it soup yet?

Taking charge of MIT-developed technology, Tom Laffey of the EPCglobal consortium says real standards are about to emerge to govern radio frequency identification technology.

When the Massachusetts Institute of Technology ended its four-year collaboration last fall with dozens of blue-chip companies researching radio frequency identification technology, the group transferred much of its expertise to an emerging standards body known as EPCglobal.

Working with the results--and many of the same people from MIT's Auto-ID Center--EPCglobal is now pushing to create RFID standards using existing communications network infrastructure and the electronic product code, a system for assigning unique identification numbers to items.

Tom Laffey, vice president at integration software maker Tibco and chairman of EPCglobal's Software Action Group, has been with the standards effort since it was at MIT and has seen the hype surrounding RFID reach new levels over the past year. Laffey recently spoke with CNET regarding his views on EPCglobal and the technology's rate of progress.

Q: The one criticism I've heard regarding the EPCglobal effort is that it has yet to publish its standard for the tech specifications to be used in RFID tags. Do you feel this is a valid complaint, and will we see a tag standard published soon?
A: It's a valid concern and, of course, it's both a software and hardware issue. I was at some of the meetings on this recently, and they're about 99 percent of the way there on at least a 96-bit tag standard. That needs to be settled because a lot of other standards will be driven by it. One of the delays has been in making sure that the tag standard is going to work outside the retail, consumer product goods area, because airlines, the Department of Defense and a lot of other kinds of companies outside of (consumer product goods) will want to use it.

The group has basically been in agreement on the standard guidelines for tags for five months. But as things were turned over from Auto-ID Center to EPCglobal, it became a little more painful. From my point of view, it's done and there's nothing that should keep the people who want to make tags from starting production, as I don't expect any major changes.

Can you talk about why the transition from Auto-ID Center made it harder to establish the standard?
Part of it is just the logistics in the move. There was a long delay related to how we might treat intellectual property issues. A number of the companies involved had concerns about that. After we resolved those issues, things picked back up.

There's nothing that should keep the people who want to make tags from starting production.
Work stopped in November and got going again in the middle of February, and it's still a little slow because not all of the companies have moved over from Auto-ID completely, because they wanted their lawyers to have time to review the new intellectual property policies. But I don't think we lost anyone officially. There are some technology companies like Microsoft, IBM and Oracle involved. And if they've gotten past the intellectual property issues, as they have, the problem is pretty much behind us.

What is the next standard you think EPCglobal will undertake?
I've been working more on the software side, but there's been a lot of heated discussion around reader standards, and that's an important one. People need to get to a place where they can interchange their tags and readers, because that's going to help drive down the costs and simplify quite a few things, including integration.

If you look at the PC market and how the Wintel architecture affected the whole ecosystem around the PC, how it drove down costs and increased adoption, that's the sort of direction most people need to move forward. I think the tag and reader standards are equivalent. There will be a lot of important software standards, but you have to have the basics in place. Once we can get the protocols down on how communication between the tags and readers works, I think we'll have achieved something substantial.

What are the pressing software standards?
There are different layers that need to be addressed. Once you get past the reader, you need to address the data acquisition and filtering layer. That's important, but not nearly as important as the reader/tag standard. And once you can filter and collect data in a standard way, you can look at how you capture and store the information, and make it available to other enterprise applications. We're working on those and they're important.

A recent report from Forrester Research indicated that a lot of suppliers are struggling to meet Wal-Mart's mandate. Even though Wal-Mart disputed the results of that research, do you think the mandates that are currently out there are fair?
Well, with Wal-Mart, a lot of companies saw the mandate but didn't take it seriously because Wal-Mart has gone through this sort of thing in the past around other issues and has tended to change terms as they've moved along. This is less about RFID than it is about how they train their suppliers to react to this sort of mandate. The suppliers know that Wal-Mart will eventually draw a line in the sand, but they expect the deadline to slip a few times before they absolutely need to react.

What Forrester found was that for many of the suppliers in Wal-Mart's program, the upfront investment to launch RFID still outweighs the technology's potential short-term benefits. Do you think that there is enough of a business case to encourage companies to work with the technology today?
There are two answers. The first is that if Wal-Mart has it as a mandate to do business with them, they can't afford to ignore it. The second, and it's funny, is the group that is pressing the hardest on RFID right now is (consumer product goods) and retail.

I would argue that there are other industries that could see return on investment from the technology a lot sooner. I think the suppliers have a perfectly legitimate gripe in the sense that there is a large upfront cost and their (return on investment) is still fuzzy. For pharmaceuticals, airlines, (Department of Defense) and others, there's a faster turnaround. With (consumer product goods), the cost of goods is so low and the scale of the project is so high, it's a lot more difficult to make a case.

Do you think this reality could potentially sully the reputation of RFID and any potential benefits it can offer in the long term?
I think that's a very real possibility, as we've probably peaked on the hype curve with RFID, and now we're heading into a stage where there's some disillusionment.

I think if there weren't the mandates out there, the market would be developing very differently.
We've had a lot of Tibco customers ask us about (return on investment), and we have to be honest about the upfront costs. Wal-Mart is telling people to run out and buy a lot of equipment they wouldn't have bought on their own. I think if there weren't the mandates out there, the market would be developing very differently. I don't think (consumer product goods) and retail would be looking that hard at RFID, and you'd see leading-edge implementations in some of the other industries I've mentioned.

What about the other mandates, from the Department of Defense, Target and Metro Group? Are these fair?
Some of the other mandates, as you dig into them, are very reasonable. They're not as comprehensive as they might appear. It's only a certain portion of the goods at certain warehouses that are involved, so I think for the most part the mandates are reasonable. No one is expecting to flip a switch and have everything working. The mandates are out there to help people do the work around RFID in phases, and I think that will work.

People have observed that EPCglobal is different than other standards efforts due to the participation of high-profile end users, such as Wal-Mart, as well as technology providers. How has this changed the standards creation and review process?
I've been involved in some other standards bodies and what's most exciting about Auto-ID and now EPCglobal is that it has been driven largely by the end users of the technology.

At all of the meetings, there are representatives from these companies pushing RFID. Tibco is also working on standards efforts around (Extensible Markup Language) and Web services, for instance. Those are being driven by the academics and the technology providers, and there's a very different tone.

How does it differ specifically from these other standards groups?
EPCglobal is all about getting things done, proving (return on investment) and establishing real standards, not just floating ideas. The end users are able to very aggressively make the trade-offs that are needed to move forward. Typically, standards groups get broken down into committees, which hold lots of meetings. Those groups move very slowly. And when you're done, you often get standards that are too complicated and too big in scope because they're trying to keep everybody involved happy. When that starts to happen at EPCglobal, we have customers there to give us a boot and help keep everything focused. Is there any feeling among the group that things are potentially moving too quickly in general, and that this could hurt RFID adoption in the long run?
I don't think so. There's a lot of discussion about why we're doing RFID now, whereas with these other standards efforts, there isn't any discussion about the economics of the business. In that respect it's very different from other bodies that I'm familiar with. It's different because EPCglobal is all about completing the standard. A lot of other groups exist purely because people like to get involved with standards for the sake of doing so; they have no incentive to complete the standard or get it out to the market.

With EPCglobal, especially for the end users, it's all about return on investment. For technology companies, like Tibco, it's sort of a disruptive technology that will allow us to build some revenue and get a sense of some new markets that we haven't been involved with in the past. I don't think anyone involved is wondering why we need to do this now or whether or not getting this standards work done will really help RFID development.

Can you give us a ballpark schedule of when you think the group will publish the various standards?
By the middle of this year, you'll have some important standards. And by the end of the year, there will be more. I think that within the next year and a half, you'll see general agreement on most of the necessary guidelines. There will be other standards we need to decide, but we should be able to do in several years' time what it took the bar code industry a decade to do.