Getting IT Right

By Bob Violino

RFID systems will present some unique challenges for corporate IT departments.

March 31, 2003 - In the late Nineties, when I was the managing editor of InformationWeek, we ran a steady stream of horror stories about ERP implementations gone awry. For those not steeped in IT lingo, ERP stands for enterprise resource planning, and it is the software that links companies internal operations, sort of like a corporate operating system. It's big and complex, and when the system installation is not managed properly, the multi-million-dollar software goes from being a major advantage to a major problem. In one famous case, Hershey Foods blamed a bungled ERP implementation for its failure to get enough candy to stores for Halloween.

As we were putting together the latest installment of our Special Report, Low-Cost RFID: The Way Forward, it struck me that radio frequency identification might well be the IT world's next ERP. The companies that get it right will get a big advantage. And the companies that get it wrong will be explaining to Wall Street why they missed their numbers for the third quarter in a row.

There are some big differences, of course. ERP was mainly software-related, where RFID is software and hardware. Another difference is ERP acts as the brains of the company. RFID is the nervous system. Since the core ERP systems are in place, companies can simple rely on their existing data-capture systems in they have problems installing an RFID system. But the more I talk to people about the IT infrastructure needed to realize the long-term vision of real-time operations based on RFID data, the more convinced I am that it's a huge challenge.

This week, our Special Report, RFID and the IT Challenge, examines the short-, medium- and long-term implications RFID on IT. The issues are significant right from the start. How do you choose which technology to deploy? How do you integrate smaller projects with existing systems. And how do you ensure that your architecture will be able to scale as more and more objects are tagged?

The issuies are complex and the answers will be different for each company depending on its operations and industry. Given the complexity involved, I frankly don't understand why many large companies are adopting a wait and see attitude. I hear many people say RFID is not ready for prime time, that the performance still isn't there and the price is still too high. Clearly, Benetton, Marks & Spencer and others realize that if you start early, you can get a cost advantage now.

Even if you believe that RFID adoption in the supply chain is two or three years away, why wait until then to begin figuring out how you will use it? Peter Abell, director of research for AMR Research's retail practice, estimates that it will take most companies a year to get funding for a pilot and then get their hardware and software systems ready to launch the pilot. Starting today leaves time to work deliberately, solve problems as they arrive and learn by trial and error.

The learning curve is steep. Even for a basic pilot, you need to examine different types of tags and readers, understand how RFID works in your unique environment, do some integration with your software systems, and upgrade some hardware. And you have to think about the impact on your business processes. There are some good systems integrators out there that can help you deal with some of these issues, but there aren't a whole lot of them.

It seems clear that many companies are doomed to repeat the mistakes some made in the 1990s. As the thought leaders came to understand the value of ERP systems and implemented them cautiously, other companies watched and waited. Then, when it was clear they could wait no longer, they rushed to install the systems and ran into problems. Rushing to install a corporate-wide RFID infrastructure in an effort to catch up is an invitation disaster. If I were an investor, I'd be selling stock in any company that says they are not looking at RFID because it's still a few years away.

Mark Roberti is the Editor of RFID Journal. If you would like to comment on this article, send e-mail to