RFID Labels for Package Printing

Editor’s note: RFID is running hot and cold in the North American packaging market. It’s a technology solution still in its early development, with many issues, risks, and potential rewards. In this column, Peter Harrop, CEO of IDTechEx, provides a glimpse of some of the behind-the-scenes factors that will impact future implementation of RFID. IDTechEx is a knowledge-based consultancy company that provides research and analysis on RFID, printed and organic electronics, and smart packaging. It will host the RFID Smart Labels USA 2008 conference in Boston February 20-21 (www.idtechex.com/USA).

It’s like Yogi Berra once said, “It’s Déjà Vu all over again.” Today, 85 percent of bar codes are now printed directly onto products—mainly packages—with no label needed, though there was once a large market in bar code labels. Ironically, the high-tech industries are the last to change, but even the electronics companies are finally looking at getting rid of the bar code labels on the back of their laptops and packaging and printing them directly on the products.

All that is being repeated with RFID a few decades later. RFID has started to replace some bar codes, but it can also be used in place of more sophisticated devices such as anti-theft tags and other items—all in one label. However, while anti-theft tags don’t use silicon chips, 98 percent of RFID labels currently do, and the chips are the most expensive part of the label. This puts the use of RFID at a significant cost disadvantage in these applications.

However, Kovio, Inc. (Sunnyvale, Calif.) has figured out how to print the thousands of transistors required by the world’s favorite RFID specification, ISO 14443, which is used in tickets, cards, e-passports, e-driving licenses, national ID cards, and some packages in supply chains. ISO 14443 is even used by the new RFID-enabled phones for purchasing items without going to the checkout. The Kovio printed transistors will go into production in late 2008 and they promise a cost reduction of 80 percent versus the silicon chip, a very important factor for packaging applications.

If Kovio stumbles, there are about 340 organizations around the world working on printed transistors. Some of that work will even lead to self-adjusting “use-by dates,” with labels (and later direct printing on packages) that monitor time-temperature profiles so they display a printed date that changes depending on how long/how hot you left the package. These will often double as RFID tags as well.

Standard and design implications

In 1999, Massachusetts Institute of Technology (MIT) came up with the idea of establishing a numbering system large enough to cope with thousands of trillions of items that might be identified uniquely by RFID one day. As a reference, five to 10 trillion items are bar-coded every year in the world but not with unique numbers. The bar-coding systems from what is now called GS1, are typically used generically, so few numbers are needed. For example, no one can currently trace a given can of soda from its bar code, but low-cost RFID makes that objective realistic, with huge implications for supply chain efficiency, safety, and security. EPCglobal was created within GS1 to devise and administer this numbering system and associated systems, with supply chains and particularly pallet and case tagging of consumer goods as the first priority.

Unfortunately, the original MIT concept of allying this to an extremely simple RFID tag was abandoned. Far from fulfilling the dream of reducing the number of transistors on the chip to a couple thousand from five thousand or so, the EPC approach of adding many “nice-if” features has led to the label on the case needing about 72,000 transistors.

A second standard is imminent. It will cover use at the world’s favorite RFID frequency—HF (13.45 MHz)—because the first standard was at so-called UHF and no frequency is ideal for all applications (or even all packaging applications). We are told by those involved that this specification will also require a similarly large number of transistors in the chip.

This EPCglobal approach will continue to find a place in the market and it may eventually lead to simplified low-cost, high-volume versions. However, others are concerned about what they see as military complexity and expense being applied even to the tagging of a gumball. These concerned parties desire administration of number issuance and membership of governing organizations at low-cost. That has led to seven East Asian countries trialling the U-Code system originated by Tokyo University. This system is aimed at very simple tags, very high-volume use, and very low cost, with less emphasis on just supply chains and what they see as the prioritization of the needs of Western retailers and CPG manufacturers.

Another low-cost approach is to have a proprietary design of a chip, such as the one used by Marks and Spencer, whose suppliers tagged more than 100 million items of apparel in 2007 and will soon tag all 350 million items yearly. This will provide what we believe is a payback of less than one year from reducing stockouts alone. That disposable tag only needs a few thousand transistors, so it is clearly an achievable objective. pP