CNET también está disponible en español.

Ir a español

Don't show this again


RFID revolution: Are we close?

IBM's Rainer Kerth says the key will be to separate practical applications from the technology's gee-whiz appeal.

When it comes to radio frequency identification technology, the conventional wisdom is that it will certainly revolutionize the way manufacturers, distributors and retailers track products and inventory.

But figuring out details of how this emerging technology should progress and get used remains a source of debate. The issues range from safeguarding data the tiny chips transmit to managing the reams of data RFID readers gather.

That's enough to boggle most minds, including that of Rainer Kerth, the architect of IBM's middleware efforts around RFID technology. Kerth recently spoke with CNET about the potential--and the potential misconceptions--surrounding this suddenly hot technology.

Q: Is RFID being overhyped?
A: I would say that there is a lot of excitement in the industry around RFID. I'm not sure if there is the typical level of overhype.

If you install RFID and continue to do what you've been doing the last five years, then you probably will not be taking full advantage of the technology.
There's obviously a lot of potential in RFID. People are a bit focused on its usage potential as opposed to how to make it really happen. It is fairly clear that RFID has the potential to revolutionize many different aspects of the existing IT infrastructure. But it's also about evolving that infrastructure to the point of supporting RFID.

When does RFID truly begin delivering all of these major benefits people are talking about?
It's actually not that far away. There are a lot of companies working under the Wal-Mart Stores and U.S. Department of Defense mandates. You have Metro Group in Germany making announcements and people in the United Kingdom working on it. Pretty much everyone in the industry is working on delivering some level of benefit around RFID in 2004 or early 2005.

Is that work going to deliver the ultimate benefits of RFID? Probably not; it's going to require multiyear efforts to take full advantage of the technology, but you can't run before you walk. I'd expect that it will be more like over the next 18 months that you see some reasonable information technology functionality to support RFID. It won't be a short-term engagement.

Are retailers better equipped to take advantage of RFID than others?
Retail and consumer packaged goods (CPGs) clearly have a leg up over others at the moment. These companies have been pretty active in the definition of RFID's base technology, and that's where they have an advantage. It is applicable to other industries, but for whatever reasons, those others haven't yet really reached the point of maturity regarding RFID that retailers and CPG companies are expressing.

Which companies have the best idea of the demands put on data by RFID?
The retailers have rich data management strategies and will clearly use them to add RFID. However, I don't think that people have reached the stage where they need to make specific implementation decisions around data management and RFID. This is something people are trying to get their arms around right now, and there are a lot of variables. What amount of data do you actually get? How do you want to present it? How long do you want to keep it? How do you relate it back to existing systems? It's too early to say that anyone, even in retail or CPG markets, has a significant lead over others.

Which standards will be the most important to RFID, and how critical will they be to the overall development of this technology?
There are several important standards that address RFID functionality, specifically radio functionality, which will be particularly crucial. RFID technology will clearly be a worldwide issue, and there must be some way to make it compatible on a worldwide level, which would be a problem right now. Radio regulations in various countries remain different, but those problems are being addressed as we speak.

What nagging issue regarding RFID has been most frequently overlooked?
That has to be the general topic of systems integration, or taking data from the RFID infrastructure and typing it back into the existing IT infrastructure...There are certainly people who know about this and may be talking about it internally, but it hasn't been documented as a significant challenge.

Are there other obstacles for RFID that you feel are not being given enough attention?
When you want to deploy RFID to hundreds of distribution centers and thousands of store locations, at some point in time, you might end up with a significant headache, if you don't design for management up front. You need to be able to monitor this environment so that you can intervene wherever some issues occur. You need to be able to get to a single system view of the overall RFID infrastructure.

How should end users approach the RFID-related systems integration challenge?
Overall, it is not fundamentally different from other similar challenges you find in other business scenarios. You have to have a multitiered approach, starting with an assessment around what specific problem you should be addressing and what the return on investment in doing so might be. From there, you dive down into the technology pieces.

Where do you think most companies interested in building RFID capabilities stand right now?
Most companies are in the process of really understanding the RFID base technology--readers, tags, data formats on the tags and the implication of having all this technology in a warehouse environment. It's about testing and exploring the impact of different materials on the RF behavior; how fluids or metals might interact with that. They're looking at where you put the tags on a given pallet or on a given case so that you get optimal read rates from those tags. The focus so far is really at the RF level and understanding how radio frequency could be used to optimize the supply chain.

What sort of IT issues will these companies be looking to address first?
At this point, I'm not sure that anyone has identified the most significant pain points. Over time, the most critical issue will be to define the right business logic on top of the base capability RFID provides and define that logic in such a way that it actually delivers business value.

Can you give an example?
Suppose that you have a business case that says you want to be able to feed information to the person at the loading-dock door who has just received a shipment. That, in itself, could be a relatively new piece of functionality, depending on the maturity of your existing operations. But it could also be new functionality that demands a process change. From a technology perspective, the challenge is to not only gather the data, feed it up into the IT infrastructure and process it there--but also to feed it to the person at the dock door in a real-time mode so that they can make appropriate decisions and essentially become more productive over time.

There is some level of new process that must be defined to really use the power of RFID. If you install RFID and continue to do what you've been doing the last five years, then you probably will not be taking full advantage of the technology.