Intermec, the Good, the Bad, and the Ugly of RFID

Scott Medford, vice president of RFID for Intermec, and his sidekick, Bob Eckles, Industry Marketing Director for Industrial Markets, came in to brief Putman editors on the state of RFID today. In my normal, shy and retiring fashion, I asked Scott, right off the bat, to explain the patent and lawsuit situation, and how it might impact adoption of RFID in the process industries and beyond. I think he was surprised by such a hardball first question, but he gamely explained Intermec's history and intellectual property in the area, with patents dating back to 1986, hundreds of them, and how they intend to handle licensing. He reminded me of the press release Intermec sent out last week, describing a new licensing program called Rapid Start, beginning June 1. This is a 90 day program with simplified and less costly licensing to the technology suite, not just individual patents, and Medford believes that many of the IP difficulties that are perceived to cloud the future of RFID will perhaps be taken care of by this program. It is certainly a much more liberal licensing program than most, and Medford asserted that their intent is to grow the market, and not to protect their market share at the expense of growth. I then asked him to comment on the Symbol/Matrix issue. You may or may not recall that Intermec sued Matrix for intellectual property problems ("It could almost have been IP theft," Medford noted, "because some of the engineers at Matrix had worked at Intermec and took their data with them."). Symbol became party to the suit when it acquired Matrix after the suit had been filed. Eventually, Scott thought the lawsuits shoujld settle down to background noise, and then be settled, but this would likely stall the Consumer Packaged Goods market for a while. He felt that industrial applications, including those in batch and other types of process industries will keep growing despite the suits. Next, we talked about the fabled and mythical 5-cent chip. "Nothing for less than 10-cents for at least the next two years," Medford replied. Bob Eckles piped up and pointed out that what is happening is a convergence of all the AIDC technologies, like RFID, barcoding, and so forth, as we try to come to grips with how this kind of technology will handle improvements to ROI. In fact, Medford said that it was his mission and that of his RFID group, "to work ourselves out of a job within the next two years," as the media hype over RFID dies down and real applications get underway. It isn't about RF, and it isn't about technology, Eckles pointed out. "Manufacturing will become event driven instread of transaction driven using the ID technologies." This will enable realtime workflow analysis and control, where before there was a big black box labeled "plant floor" with a goesinta labeled raw material, and a goesouta labeled finished product. What happened inside the magic box was unknown and unknowable. Terra incognita is no longer practical if you want to produce the kinds of production innovation that will keep the US and the EC ahead of the rest of the world. Intermec is "all about data-- fast, reliable data," Eckles said. After talking about their quest for ATEX certification of the intrinsically safe handheld terminal running Windows CE (or whatever the new name for it is) that they showed us, we started to talk about actually having applications where actual users are actually using the products in the process (and machine builder) markets. They promised us some case studies, and I hope to see them soon. After all, we can only talk about how cool a technology might be for a while and then we have to "show the money." Our readers want to know if this is real stuff, or vaporware. The real issue, Medford and Eckles believe, is that the future is about "putting information where the people are." This is diametrically opposed to the monolithization of data management by the traditional ERP giants like SAP. Whether you subscribe to the URL path I talked about in January (see the January page in this blog for the meeting at Rockwell) or to the database path (where you put all the data right on the chip), the key is to produce a realtime product genealogy for each item of WIP. This is a real big change in the theory of manufacturing and data analysis. I mentioned to them that Emerson has already started the use of "intelligent agents" in their Machinery Health Monitor, and Eckles noted that he would rather see "report by exception" than the tons of unmanageable data managers now get. A view of the future, or pie in the sky? We may not have that long to wait to see. Comments? --Walt Boyes