Jul 22, 2002July 22, 2002 - Back in the late 1960s, Alan Haberman was executive VP and later CEO of First National Stores, a chain of grocery outlets in New England. He was dealing with a lot of the same problems that were plaguing the entire U.S. grocery industry at the time. Inflation was driving costs up. The conversion to supermarkets from small stores was just about finished, and the competition for volume was driving down margins. Productivity was stagnant, and it was getting harder and harder to make a profit.
Desperate to turn things around, talk among leaders in the grocery business turned to how the entire industry could be changed to reduce costs across the board. One obvious and frustrating issue was the failure, after years of discussion between retailers and grocery manufacturers, to agree to a standard code to identify grocery products. It was a simple concept that could bring efficiencies to producers, distributors, logistics providers and retailers.
The various trade associations representing food retailers and manufacturers met in late 1969 determined to make the product code happen. Four months later, they gathered leaders of both segments of the grocery industry at the Mayflower Hotel in Washington, D.C., and decided to form a small committee of the respected CEO to figure out how to develop a code. The group would have to overcome political differences and special interests to create something that would benefit everyone.
The Ad Hoc Committee, as it became known, was composed of five executives from grocery manufacturers and five from the retail distributor associations. After the first few meetings, they decided the Universal Product Code, or UPC, would have five digits to identify the manufacturer (assigned by an industry control group) and five digits chosen by the manufacturer to represent the individual stock-keeping unit (SKU).
On March 31, 1971, the Ad Hoc Committee created the Symbol Selection Subcommittee and chose Haberman to chair it. The group's task was to find a machine-readable version of the Universal Product Code. It had to pick the technology for an automated front end and prove that it was economically viable for both individual companies and for the retail and manufacturer industries as a whole.
Two years, dozens of meetings and lots of arguments later, the subcommittee settled on the bar code. In April 1973, the Ad Hoc Committee presented the UPC system to the industry. Adoption was by no means a given, so the industry created the Uniform Code Council (UCC) to manage the transfer of the technology to industry. The council would be staffed by key players in the ad hoc effort and people nominated by the various trade associations. Haberman was among those chose to sit on the council and remains on it to this day.
During the symbol subcommittee's investigation into identification technologies, it consulted experts appointed by MIT to advise them. The scientists said that the bar system developed was state of the art, but also that in 25 years, a new technology would likely come along to replace it. In 1997, as the 25th anniversary of the bar code approached, Haberman was asked by the UCC Board of Governors to search out research teams that could help the UCC and EAN International (the UCC's European counterpart) look at new technologies that offered advantages to current systems.
He spent two years traveling around to universities, looking at their capabilities and at technologies being researched. By mid-1999, it was clear to him that RFID was the right technology and MIT was the place to do the research. Three faculty members there, Sanjay Sarma, David Brock and Sunny Sui, were already exploring the idea of using low-cost RFID tags and a related IT infrastructure to identify individual products and things. The central idea was to store only a unique serial number on each tag and store related information in databases.
Procter & Gamble, was active in the MIT Media Lab, and its representative, Kevin Ashton, introduced Haberman to the three researchers. In September 1999, UCC and EAN International, agreed to back the creation of a lab at MIT to research the potential for using RFID or perhaps other technologies to identify individual items. P&G and Gillette agreed to match the UCC/EAN contribution. The MIT Mechanical Engineering Department added to the initial funding, bringing the total to $1 million.
The Auto-ID Center was announced at the Smithsonian on Sept. 30, 1999, at a celebration organized for the 25th anniversary of the Uniform Code Council. The center was formally established the next day. Ashton became director. Sarma was named technical director. And Alan Haberman was elected chairman of the Board of Overseers. He remains on the executive committee and is intimately involved with the effort to create the Electronic Product Code (EPC), a system for tracking individual items using low-cost RFID technology. He recently spoke to RFID Journal Editor Mark Roberti. Click on the link to read excerpts of their conversation.
RFID Journal: There have been numerous stories lately proclaiming the death of the bar code. Do you think RFID will one day replace bar codes?
Haberman: Bar coding is a very mature market. There's a gigantic built in user base. There will be a migration towards RFID. But there is a great deal to learn [about RFID] and a great deal still to do. So they are going to live side by side for many years. I see RFID eventually taking over the identification world. Bar codes will be used mainly by smaller companies and for individual projects, where the capabilities of the EPC or GTag [a UCC/EAN-supported standard for RFID tags] aren't required. They will be mainly closed systems that do not need interoperability. But it will be a heck of along time, if ever, before bar codes are marginalized.
RFID Journal: There are a lot of people who say the Auto-ID Center's system isn't viable. Were there a lot of naysayers when you were developing the bar code?
Haberman: The kind of excitement that you are seeing now was happening back then, in the early 1970s. At the same time, the electronic cash register was just starting to show up. People were starting to put numbers on items. Instead of reading them by machine, the idea was to type in a five or six digit code instead of the price. It gave them a lot of information that they didn't have before, but it was very slow, and there were no direct efficiencies or hard savings at the front end. The unions questioned automating the checkouts. Consumer groups fought us. A lot of people thought you would never get everyone to print the codes and that you would never be able to manage it as an open system.
RFID Journal: Was the cost of lasers seen as a huge obstacle?
Haberman: Part of my committee's job was to determine what the cost of the system would be and whether there would be a valid return on it. We had to show a return for the entire industry. We did that through hard savings -- labor savings and some savings in shrink. We took very little soft-cost savings into account.
We were looking at something over $100,000 to outfit an average store in the early 1970s. We were sure that would drop to $25,000 to $50,000 over time. We were able to go ahead because there was a return even at the high point of cost. The early adopters would get a valid return on investment, and you didn't have to go too long before volume production savings would reduce the cost to something very reasonable. Then, the return on investment would increase dramatically.
RFID Journal: Are there any major differences between the effort to create the bar code and what's happening now with RFID?
Haberman: One difference is our effort was only in the United States. EAN didn't come around until four or five years after the UCC. Now, we have a global supply chain, and the effort comprises the whole world. Another difference is that the Auto-ID Center is building on the structure that UCC and EAN put together --and computers have developed beyond our fondest dreams. We didn't have anything to build on.
And the Electronic Product Code doesn't just apply to the grocery industry. It was conceived to apply to all industries and all areas where identification is necessary, including the distribution system, logistics and marketing. The bar code started with the marketing system, and it took years to get it into distribution. Now, we're doing it all at once.
Another important difference is that our hard cost savings at the checkout required little integration into or change in surrounding systems and technology. To get other benefits, however, you had to change the way you did business. We have always been disappointed by how little of those gains were earned. Will industry make the systemic changes required by the EPC? I worry a lot about that.
RFID Journal: The UCC is backing the GTag standard and the EPC. Is there a conflict between the two?
Haberman: No. The GTag proposed standard says it will be compliant with the EPC and the EPC proposed designs would be compliant with GTag. All GTag is doing is identifying the SKU, like you do today with bar codes. It doesn't change the whole infrastructure that you have to deal with, so it gives you the opportunity to move step by step into EPC, where every item is identified. People don't understand that EPC, as envisioned, requires a tremendous change in information and system management to realize the huge potential benefits. You identify only the item, and everything else is done by reference to databases connected via the Internet. That's a tremendous change, and reality demands that it be done in intelligent and planned steps over time.
RFID Journal: There are a lot of people who say they'd rather stick with bar codes because they are cheaper. What's your response?
Haberman: There are always those who want to stay with the old system. There are lots of people who won't spend the time to determine the return on investment. An RFID tag may never approach the cost of a bar code printed on a label, but the returns will be terrific. There were lots people who went out of business because they didn't keep up with bar code technology. The same thing will happen again. But EAN and UCC have determined to help industry coalesce around RFID technology. That helps move things along and leads to an open, widely accepted system.
RFID Journal: Is EPC adoption inevitable?
Haberman: There is very little question in my mind that this is the future. Whether it is exactly in this format, I don't know. But it is going to happen. It is a question of time, energy and management. It has to be managed, so that when problems arise - and they will arise - they get solved. There has to be an organization turning the center's recommendations into standards, doing education, and mounting additional research. The center and its industry members are working now on the shape of the organization that will take on the task.
RFID Journal: What could derail the effort?
Haberman: A couple of things. Some vendors marketing closed systems for years have fed industry false information about what RFID is capable of. And RFID suffers what I like to call the "future magnet" syndrome. Present a futuristic idea that is grounded in reality, and the press and talking heads looking to impress people just can't resist attaching to it every farfetched dream they can describe. They ignore cost and effort and time, the inertia of existing investments and habits. There are bound to be tremendous disappointments if people don't soon learn what the limitations are. The laws of physics need to be understood. Radio waves don't go through metal. They are absorbed by water at some frequencies. You have to work around these shortcomings. It will sink the system if people try RFID and expect more than it is ever capable of giving them. It is a possibility today. False expectations could kill us.
There will be companies that don't think the total system through and buy a vision they have. They may try simpler, niche implementations and find there are limitations. These guys will not spend the time and energy to solve some of the engineering problems, like where you place the readers, how you shield RFID equipment from other RF devices and so on. These fiddlers will run away having poisoned the environment.
And the reverse is also a serious problem. Some companies that would have tried RFID for important projects will stand on the sidelines until the expected cheap chip arrives or until others work out all the problems. In the process, the young and fragile inventory of supplier companies will starve and disappear. Choice will be eliminated; imaginative minds will go elsewhere. Is it any wonder that EAN/UPC are bringing GTag to market?
RFID Journal: Are you optimistic about the future of RFID and the EPC?
Haberman: Sure. A year ago, there wasn't anyone who thought we could get the price of tags down to 25 cents, no less a penny. Now you can buy 25-cent chips if you try. It will get lower because supplier companies are doing what they see the market requires. The Auto-ID Center, by presenting its vision and working towards spelling out the details, is already shaping the market. Its final recommendations -- and note, they will only be recommendations because the center does not write standards -- will be turned over to industry. When and if industry turns these recommendations into global standards, as it learned how to do with the UPC, the market will explode.