Identifying RFID's Biggest Threats

Violating the Concept That the UII Must Be Monomorphic
The second threat—violating the notion that the UII must be monomorphic—is just as insidious as the first. Together, they reduce EPC Gen 2 to a state worse than where the industry was prior to Gen 2.

The basic problem is that when the first EPC specifications were published, very little was said about programming user memory. Many early adopters of Gen 2 technology relied on solutions that did not require any user memory—partly because in the early days, no chip supplier sold tags that contained user memory on-board. Early suppliers of Gen 2 chips offered only 64 bits (later upped to 96 bits) for the UII/EPC Memory Bank (MB01), and nothing else for additional memory (User Memory—MB11).

Recently, chip suppliers have begun releasing chip products with small amounts of user memory (up to 512 bits). Some companies feel that adding incremental amounts of information to MB01 is the fastest way to access "what they want." The correct answer is to place the intended and accepted construction of UII/EPC into MB01, and all other information in MB11. If companies begin to place "user information" (data intended to be placed into MB11) in MB01, that would create a scenario similar to the "hello" and its echo, versus the longer conversation described earlier. Readers would begin to become confounded by the additional information required to simply say hello. And we aren't considering the effects real-world RF interference would play in the longer message lengths. RF interference has always been the bane of RFID—more so today with the proliferation of RF-based tools in the workplace and throughout the supply chain: When RF interference is present, the message may have to be sent more than once, and can take longer.

Again, history teaches us something about user memory. In the first releases for LF, HF and UHF tags, each had TID only, with no extra memory. Within two to four years, however, LF, HF and UHF tags each offered up to 2 kilobits of memory. Unfortunately, that message was lost in the evolution of Gen 2 technology. Chip suppliers needed to know how memory to offer. Since the retail sector represented more that 80 percent of early UHF RFID Gen 2 adoption, and retail had no user memory requirement at that time, chip suppliers lacked a clear mandate and began adding user memory to chips only incrementally. Retail is unlike the automotive or aerospace sectors, in which user memory has been historically employed and 2 kilobits has historical meaning. To no surprise, the 2-kilobit milestone is now upon us.

The fundamental problem lies in the intermediate-sized tags, where the tease is to just add some small extra piece of information into MB01 rather than to follow the standard and historical practice and place 100 percent of user data (small or large) in the area reserved for user data—namely, MB11 (and never MB01).

Now imagine that interrogators would need to not only identify the UII, but also capture random additional information as well. This thinking is basically proprietary in nature, undermines universal exchange and reduces the Gen 2 standard to a convenient interface and protocol for proprietary exchanges.

What Needs to Be Done
With this in mind, we need to honor standards and the need to maintain monomorphic UIIs, and not try to deviate from them for shortsighted benefit. Doing so will be key to the near-term assurance of interchangeability and long-term access to the "Internet of things."

Patrick King is the current leader for global electronics strategies for 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), and 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 and operates 19 major manufacturing plants in 17 locations.