With 2,000 retail companies, 2,000 suppliers, and more than 90,000 convenience stores in its membership, the National Association of Convenience Stores (NACS) is a large, diverse organization. John Hervey, NACS's CTO, explains Project JXTA's role in the organization's ongoing technology standards program.

Boardroom Minutes: Efficiency seems to be important to your members. How long have you been working to improve efficiency through technology solutions?

John Hervey: We started the technology standards project in November 1995 and have what we call four pillars as its foundation. The first pillar is electronic business-to-business, or EB-to-B—originally EDI.

We didn't get much traction for traditional EDI, but when XML came along, we took our existing work and put it into XML format. Today I think we have more active implementations of XML EB-to-B than any other retail segment. Every day, for example, Shell Oil sends out approximately 17,000 price change notifications in our format, and we have bills of lading and other documentation in the system.

The second pillar addresses payment systems. This deals with the message format that flows between a credit card processing terminal at a retailer location and the host credit card processor. That message format, believe it or not, is different for every processor. As a result, when you go to a gas station and use pay-at-the-pump, the message format differs among the major companies. To help our members, we've adopted an international ISO standard for a standard message format from the credit card terminal to the host. Several of the oil companies worked on that with us and will be making that change as they update their software.

The third pillar addresses point-of-sale (POS) and back-office integration. In our industry, retailers like to be able to purchase a POS from one vendor and a back-office system from a different vendor, so that they get best-of-breed solutions. This often leads to the necessity of having an interface written between disparate vendors, which is very expensive. So our effort here is to develop a common standard for interchange between POS and back-office systems.

The fourth pillar addresses device integration and here's where we get into Project JXTA. We're developing what we call the NACS Open Site Architecture. We're encouraging vendors to adopt it, and we're going to be encouraging retailers to do so through some pilot projects—actual field installations.

BM: And you're encouraging your members to adopt Project JXTA technology?

JH: Well, we're not going to say, "We want you to adopt Project JXTA technology." We're going to tell our members that we want them to adopt the Open Site Architecture, which is built on Project JXTA protocols and a couple of other things.

BM: What is the background of the Open Site Architecture?

JH: First we considered all the devices that sit on the counter in a convenience store and the fact that they don't talk to the POS system. This is, of course, a real accounting and reconciliation nightmare for a retailer. For example, if a clerk sells a money order and processes it on the money order terminal, the retailer then depends on that clerk to go over to the POS and ring it up. If this doesn't happen, the reconciliation document from the money order company will show that the money order was sold, but the retailer won't know when it was sold, who processed it, or anything else.

BM: And this is only one example... .

JH: Sure, the same thing is true with lottery tickets. Every time an online lottery ticket is sold, the retailer depends on the clerk to do it on the POS as well. So what we wanted to do was to get everything talking to the POS.

BM: What other devices are there?

JH: Well, there's the tank gauge monitor, there are fuel dispensers, there's a car wash code generator, a safe, and of course the back-office system... .

BM: So a tremendous number of inputs you need to have in your centralized communications.

JH: Yes, and you also need off-site people to be able to go to specific devices. For example, an armored car service may be hired and may want to monitor the safe, and only the safe. It needs to be able to come in and find out how much money is in the safe, and that's the only information you want that service to be able to access.

BM: So this example shows how a JXTA peer group is employed, right?

JH: Yes, that's right. We wanted to get everyone talking to one another and then be able to define exactly who has the authority to talk to which device or devices. Our objective was to create an architecture for the convenience store of the future, three to five years down the line, and define how we will get there from where we are today.

BM: How long have you been working with Project JXTA technology?

JH: We started talking with Sun in late August, early September 2002. We put a group together that developed questions for Sun. Sun's responses led us to make a decision to move forward with Project JXTA technology. Then in December, Sun produced a training seminar for us.

BM: So what are the benefits of the Open Site Architecture and Project JXTA protocols?

JH: There will be substantial savings, number one, for convenience store owners, because we think that it will be easy for them to implement solutions. They will be able to build to one standard, as opposed to diverse platforms. They will save money, because they're going to have to do only one kind of implementation in the future.

BM: Then their suppliers can compete on price and service?

JH: Right, but we think retailers are really going to benefit in other ways, such as in maintenance savings: A dispenser manufacturer in a dispenser peer group can use JXTA protocols to do some diagnostics and preventive maintenance.

Another example is today's tank gauge monitors, each of which is hooked up to its own modem, reporting leaks or high-water problems. Being able to put this function onto an Open Site Architecture network means that a tank can be part of a peer group and notify people through a pager or a cell phone. The retailer is thus able to remove a modem and a phone line that are no longer needed.

I also think fewer devices may be necessary. Things that are proprietary today and need to be stand-alone devices can, under our Open Site Architecture and JXTA protocols, perhaps be integrated into a common POS device if all they're doing is communicating some information.

BM: That should help stores handle individual customers more quickly.

JH: Exactly, and speed of service is very important in the convenience store industry.

BM: Is there an enterprise benefit to retailers that have more than one store or a chain of stores?

JH: Oh yes. Suppose I have ten stores and I want to change the price of gasoline at all of them. I can create a peer group from my home office that includes the price sign, the fueling dispensers, and the POS. And I can send out the new price to the fuel dispensers and get confirmation that the prices have been changed—and I can do that from a home office to all ten stores.

BM: It seems that in this arena, there are two security aspects: the security of the retailer's information and the privacy concerns of people's ATM cards and credit cards.

JH: There's not really a concern about ATM cards and credit cards, because all of this information is already encrypted. But we are concerned about someone trying to hack into the IP network and across to some device and do some damage that way. We think Project JXTA's security is going to protect us from that.

BM: How much technical understanding will an individual store owner or clerk need in order to work with the Open Site Architecture?

JH: If this thing is going to be successful, it's got to be simple. It has to be something that's easy to use on-site but that also provides ease of use for off-site adjustments and monitoring, and so on.

BM: Are you already thinking ahead to further steps?

JH: This is an iterative process, and one probable next step for us involves POS/back-office integration. We want to take all the work we did in XML and then port it over. The device integration work—the Open Site Architecture—and the POS/back-office work will come together. I think that this will be a tremendous thing, because we'll have combined the benefits of both of these initiatives.