Maybe it's because this is Thanksgiving week, or that Christmas is just around the corner, but we really are feeing a bit concerned about thousands of suppliers of that multi-billion-dollar U.S. retail institution known as Wal-Mart. Now that Wal-Mart has decided to plunge head first into radio frequency ID (RFID) tagging, a lot of these suppliers are a little worried that the company may be rushing too fast into relative virgin territory as it looks to become a technology pioneer.
Wal-Mart has been working with IBM and its RFID 'tag team' for a number of months, most recently as part of a six-week pilot project designed to set the stage for the widespread deployment of passive RFID tagging systems (the kind that do not consistently transmit information, but only do so when activated by a specialized RFID reader). This pilot project literally just wrapped up a week or two ago. The next step is a more aggressive initiative that will affect 100 or so of the company's top suppliers who ship products to a centralized distribution facility located in Texas. Following this, all of Wal-Mart's suppliers will be under the gun to apply RFID tags to all of their products by January 1, 2006.
Wal-Mart suppliers are a bit concerned because the company's passionate commitment to RFID tagging has quickly evolved into a supply chain mandate, which means that if you do not comply then you may cause the firm's annoying little happy face to frown, and that
At a recent meeting on RFID tagging held at COMDEX in Las Vegas and moderated by Shoreline Research, nearly one-third of the audience of 125 or more people raised their hand to indicate they would be affected by RFID tagging initiatives in the very near future. Questions from the audience, which followed the session, came fast and furious -- especially those targeting two executives from IBM Corp., who's Global Services Division was involved in the Wal-Mart project. In fact, we could have gone on discussing the issue for another hour or so, if not for another session following close behind.
We are not usually prone to hyperbole, but it seems that Wal-Mart's RFID tagging enthusiasm may be heading directly into a condition the meteorological-inclined refer to as the 'perfect storm.' In this case, however, the storm patterns include: An improving, but fragile economy; a lot of fear, uncertainty and doubt among the suppliers and potential users of RFID tagging; and a technology that is still rife with multiple standards, evolving regulations, and relatively little testing. In fact, one of the IBMers on the COMDEX panel, who was directly involved in the Wal-Mart project, chief RFID architect Doug Martin, referred to the current RFID initiatives as a 'giant experiment in physics.'
Now, we know what you are going to say: RFID technology is not new and has been around for a while. Also, there are some very strong and well-documented examples of RFID tagging not only resulting in a significant cost savings over a very short time, but also powering some tremendous boosts in productivity -- especially in medical applications, where RFID is used to track and locate necessary supplies and medicines for doctors and nurses. You might even point out that RFID technology is not meant to replace the familiar Universal Product Code (UPC) symbols that are not slapped across every product we buy, but can be used with UPC technologies to sharpen supply chain logistics and ultimately lower the cost of goods to consumers.
Feeling a bit cocky, you may even remind us that this first wave of RFID involves limited-range passive communications technology (so, it is not as intrusive and communicative as some people would lead us to believe), and will not be used to tag individual products (but does target pallets and boxes).
All valid arguments, to be sure. However, there are still a lot of issues surrounding the technology and its widespread use, which makes us feel it, might be better to proceed with RFID a little slower and with a lot more caution than what seems to be the case with the Wal-Mart blueprint. Some of our concerns include:
- A lack of standards when it comes to what kind of information, and how much information should be stored in individual RFID tags. Right now, most tags are designed to hold up to 64 bits of information -- details like product description, inventory codes and point of origin information. However, Wal-Mart is reportedly looking to expand that capacity to 96 bits, and design tags so that they have a one-time writable memory.
- Multiple communications network infrastructures that span the licensed and unlicensed spectrum, and get even more off the scale if you look beyond the U.S. For example, most RFID tags in the U.S. operate on either 900 MHz UHF or 802.11 Wi-Fi networks, and can broadcast data up to 30 feet. Previously, these tags -- which, as we said, have been around for a while -- broadcast over a bandwidth ranging from 125 KHz to about 135.5 KHz, and could only transmit about 3 feet. Move outside the U.S. and you are likely to hit even more standards and spectrums.
- Communications problems and disruptions when dealing with multiple RFID standards and a plethora of different readers needed to talk to and communicate with these disparate tags. IBM's top RFID honchos claim there is the strong possibility of signal collision and errors when you start using different types of readers within the same area.
- Control and management of RFID tags and labels, which now cost about 50 cents to one dollar each to produce (since consist of multiple components and are manufactured using some pretty neat technology) and will number in the trillions when used by the six or seven top retailers in the world. This is a particularly troublesome issue if these tags are communicating over 802.11 Wi-Fi networks, and evolve from occasionally-transmitting passive to consistently transmitting active architectures.
So, with all of this activity, what is our advice for the retailer or systems integrator looking to get a toe-hold in RFID? For this, we defer to our friends at Zebra Technologies International, LLC., which also took part in the COMDEX panel in Las Vegas. They suggest that you:
- Consider if the RFID target is a single company that plans to keep all of its RFID activity in-house, or a large company that is looking to tag all of its suppliers and link with their RFID systems.
- Identify the proposed environment for RFID. Will it be primarily located on or near a loading dock or relatively pristine office, or will it be positioned within the bowels of a messy manufacturing area? This is important since you must consider the type of labels you want to produce.
- Weigh the cost benefits of RFID. If you are a smaller operation, then cheap UPC symbols may be a more worthwhile alternative than RFID tagging, which might cost up to $1.00 per unit.
- Specify your performance goals and aspirations. Do you want fast speeds and the ability to read multiple tags at the same time, or will your scanning requirements be relatively modest at first? What about the volume of data you want to store on each tag? The more data storage capability, the greater the price per tag.
- Finally, determine if you need a dual system (printers and encoders) that allow you not only to read tags, but to produce them on the fly. This can result in a long-term cost-savings by eliminating unnecessary and duplicate equipment.
- Be sure to scan this column for more on the evolving RFID technology story in the future!
This was first published in November 2003