May 25, 2009I believe that 100 percent of Electronic Product Code (EPC) Gen 2 RFID and ISO-18000-6C applications, beyond the most basic EPC applications, are proprietary and closed-loop. My fear is that some new clients for EPC Gen 2 and ISO-18000-6C ultrahigh-frequency (UHF) RFID applications in 2009 will wrongly believe they are buying a globally interchangeable application, when their solution is actually closed-loop.
I was duly impressed with the number of attendees and the caliber of offerings at RFID Journal LIVE! 2009, held in Orlando, Fla., on Apr. 27-29. The majority of the demonstrations were tied to critical business cases and real requirements met by those applications. The relative maturity of the industry was also evidenced by the inherent interoperability of devices—namely, chips, tags and readers. However, just because the hardware is interoperable does not mean the applications in which it is deployed are interchangeable.
Before I provide some definitions, I need to make a few points very clear. First, when RFID tags and interrogators receive EPCglobal certification, it merely means that for a certified EPC application, the tags and readers are 100 percent interchangeable and interoperable. The second point is that the number of certified EPC applications outside of retail and the U.S. Department of Defense (DOD) is small compared with the total number of emerging applications that utilize the exact same readers and tags, but for applications that do not comply with the EPCglobal and ISO-18000-6C standards. And finally, for 100 percent of applications that are EPC-based (but include additional user information) and 100 percent of non-EPC applications (including those for which Protocol Control Bit 17h is a binary 1 and the application follows ISO 15962 encodation), there is 100 percent interoperability but zero percent interchangeability today (see Identifying RFID's Biggest Threats).
The concept of interoperability is very simple to understand. At present, you can buy a cell phone from many suppliers and resellers, and you can also change service providers. This is a good example of interoperability. For quad-band GSM cell phones, we even have interchangeability. Just plug in the SIM card of the operator you want to use, and it becomes part of that network—regardless of where or how it was purchased. At RFID Journal LIVE! 2009, there were great examples of interoperability for tags and readers. The tags could be programmed and read by interrogators from many vendors, service providers and resellers, and the readers could do likewise. It was very much like the cell phone analogy, which I like because the cell phone is a wireless example not too dissimilar from RFID.
This is truly an exciting time for the auto-identification industry, and I am reminded of the early days of 2-D bar codes. At the Scantech Expo in 1989, many companies could print Data Matrix 2-D labels, and many scanners could read those labels.
If that was so good then, what is the problem now?
The issue is that both in the 2-D example and at LIVE! 2009, there was almost no interchangeability. The machine-to-machine interfaces (as in reader to tag, or imager to bar-code label) were fully interoperable, but the interrogators displayed only the raw data—in this case, the machine-based hex programming code in which the RFID tag or bar-code label's data was written. The readers did not (or could not) translate the hex code back into the original human-readable words and numbers. The exceptions were when the label and readers were matched by employing the same proprietary encodation process.
I walked around the exhibit floor with a section of tire that had an EPC unique item identifier (UII) number and user memory programmed into the EPC Gen 2 RFID tag embedded in the tire. No vendor, service supplier or exhibitor present at the event could read the actual data encoded to that tag. They could read the bit code and hex used to represent the UII number and user memory, and then report the hex and bit code, but none of them could tell me (in ASCII) the actual UII and user data written onto the tire tag.
Interchangeability is the ability to communicate interchangeably and, for our industries, essentially globally. Let me use a cell phone example one more time. The best way to describe interchangeability is to imagine that I am calling someone within the United States, and that I reach the wrong telephone number. The process is simple: I ask to speak to a particular individual, and the person on the other end replies that I must have made a mistake. This seems very simple, and works with any combination of phones and network operators, which illustrates interoperability and interchangeability in this "wrong number" exchange.
Now imagine that I get a wrong number while making a call to Europe or Asia. In such a scenario, the person answering the phone may not understand English. In this case, there is no useful communication—or, stated differently, it lacks interchangeability, even though there is interoperability.
Before I discuss the LIVE! 2009 exhibitors once more, let me take you back to the 1989 Scantech show. That was the first year a large number of vendors demonstrated label printers printing beautiful Data Matrix 2-D bar codes with a lot of data onboard. The benefit over traditional (linear) bar codes was the ability to carry a greater amount of information, and to concatenate that data to add even more. Alongside the printer displays were the latest fixed and handheld CCD imagers (CMOS was not yet popular), all of which could successfully read the 2-D Data Matrix codes. Life seemed great, until you took the label from one booth to the reader of any other booth.
None of those other readers could report anything except the label's hex code, unless that label was produced via the proprietary code used by the reader vendor. Here was a very early example of 100 percent interoperability and zero percent interchangeability. Fast-forward to 1999, and nearly 100 percent of imagers read 100 percent of 2-D labels, except in situations in which proprietary encodation was applied, as in a closed-loop application.
Now back to RFID Journal LIVE! 2009. This year, encodation (conversion of machine code to useful code or information) is at the level that 2-D bar codes were in 1989, and nearly all vendors at the event defended that status, claiming most current applications for UHF RFID are closed-loop, and that it is easier and faster to do what you and the customer both want.
The fallacy of this way of thinking is the focus on the "current project" and its short-term gain. Intuitively, what these vendors told me seems correct, and maybe even OK. Privately, several told me they want to continue using their own encodation because that gives them control of the customer and, potentially, that vertical industry. All of that may seem true, but (a) in this downturn in the global economy, there will be mergers and acquisitions, and many of those righteous vendors will be bought, sold, merged or closed—and what do you think their customers will do with their proprietary encodation? (b) a truly global client demands interchangeability, and (c) the unparsed hex format is not commonly acceptable in higher-level systems such as SAP, and will cause significant problems in a world where Web-based information exchanges are expected.
Raw hex code may be data, but to become actionable information usable in applications, it must be parsed into fields with meaningful values. A proprietary encodation system above the hex may be useful in a single company's closed-loop system, but it will not function across a complex supply chain with constantly changing lists of participants. Hex was what was demonstrated in 1989 at Scantech for 2-D bar codes, and hex was what was demonstrated at RFID Journal LIVE! 2009. Interoperability employs hex within auto-ID solutions, but the key for interchangeability is to have the higher-level standards above hex applied.
What is the answer? The solution is to employ standards. In 1999, 2-D bar codes migrated to ISO encodation. Once the complete standards were published, adoption was slow but ultimately OK. RFID must use standards to achieve the same global success for interchangeability. I told someone the other day that I was not panicking, and that person told me I should be. But I am not panicking, mostly because the ISO 18000-6C/EPC Gen 2 standard is far more mature for interoperability and less mature for interchangeability. I see rapid adoption and success with interoperability for EPC Gen 2 RFID, and I see improvement with interchangeability on the way.
The fact is that both EPCglobal and ISO have new standards being released this year, which are almost 100 percent focused on interchangeability issues. EPCglobal is finalizing version 1.5 of its EPC Tag Data Standard (TDS), while ISO plans to publish a revised version of its ISO 15962 standard, which defines rules for encoding data to RFID tags. The updated EPC and ISO standards will include information regarding programming user memory.
Additionally, several vertical industries are expected to publish ISO guidelines for programming the UII memory—that is, memory bank 01 (MB01)—as a counterpart to EPC. What's more, the Automotive Industry Action Group (AIAG) plans to issue Revision 8 to the B-11 automotive item-level standard. This revision will include 100 percent of the details to secure interchangeability within that industry for all versions of ISO and EPC encodation.
Next year's RFID Journal LIVE! conference should include many demonstrations of implementations for which interoperability and interchangeability are complete and global.
Acknowledgement: The author wishes to recognize the significant contributions made to this article by Bill Hoffman of Hoffman Systems LLC.
Patrick King is the current leader for global electronics strategies at Michelin, and the tire maker's representative to EPCglobal. He is also a member of Global AIDC 100 and AIM Global's RFID Experts Group (REG), as well as a contributor to RFID for Dummies. Dedicated to the improvement of sustainable mobility, Michelin designs, manufactures and sells tires for every type of vehicle, including airplanes, automobiles, bicycles, earthmovers, farm equipment, heavy-duty trucks, motorcycles and the space shuttle. Headquartered in Greenville, S.C., Michelin North America employs 22,600 workers and operates 19 major manufacturing plants in 17 locations.