Terminal Illness: new schemes on old POS machines

Question: What do the Mars Rover and the New Zealand Eftpos infrastructure have in common? Answer: They are both boldly going forth on the back of some very old IT.

Question: What do the Mars Rover and the New Zealand Eftpos infrastructure have in common?

Answer: They are both boldly going forth on the back of some very old IT.

In the case of the Rover, its 1970s vintage 80C85 processor need take it only across the bumpy surface of another planet. Our more than 40,000 Eftpos terminals arguably have stranger and more difficult terrain ahead of them.

The natural desire of merchants to avoid the situation in the 1980s, where as many as four different POS terminals crowded retail counters, will see new smartcard and stored value card schemes, such as Mondex and Visa Cash, layered on top of the existing infrastructure. But some vendors warn that it will also see new schemes locked into the limitations of that infrastructure.

Mondex, with Multos, and Schlumberger, with Solo, its implementation of the JavaCard API, now tout the,feaures of their smartcard OSes — both systems emphasising the ease with which, say, a new customer loyalty application can be easily and securely loaded on to cards — online, if necessary.

But adding the same application even to terminals which are “smartcard-ready” probably means going on-site tothe software. Assuming, that is, that there’s room for the SAM. Many standard terminals have room for only four SAMs.

If, as Mondex proposes, you’re planning to do public key cryptography on the spot, you’ll be wanting to add some memory (at $250 a megabyte) and hope the 8-bit terminal is up to the job. And you’d better keep your loyalty scheme simple, because the customer isn’t going to want to be standing around poking at a keypad making choices for long.

“It’s absurd trying to to add this new technology to that old framework,” declares Clive Cooper-Smith of the Advantage Group, which sells Hypercom terminal equipment. “That point-of-sale network framework that’s out there now has been out there for 12 years and it hasn’t changed. The only thing that’s changed in Eftpos terminals is that more memory’s been added.”

Geoff Gander of the Australian-based firm Intellect agrees that most POS terminals are “archaic” in today’s terms, and takes up Cooper-Smith’s complaint about the difficulty of introducing new applications to the system.

“Let’s say you have a device which is capable of running multiple schemes,” says Gander. “You say, I now want to install a scheme. First, you have to download the application over the wire. Then you’ve got to open the device and implant the SAM into the SAM holder, so you’ve got to be capable of having multiple SAMs. Assuming you do both of those, you’ve got another scheme running on your device. But a lot of people still have no idea how complicated that really is.”

Cooper-Smith’s competitors imply that he and Hypercom are keen to play up the difficulties of upgrading terminals because they have more problems with upgrades than other terminal vendors, but Cooper-Smith has been calling for some time for a switch to a “client-server” system, where applications live on a server, rather than in a SAM.

“In a perfect world, yes, that would be the case, but we don’t live in a perfect world,” says Ian Bailey, managing director of Keycorp NZ, whose company helped develop the SmartTaxi SVC scheme running in Wellington. “There’s already 40-odd thousand terminals out there, of which the majority are either owned or rented on a fixed-term contract. To go to those people and say, I’m sorry, you’ve got to throw away your piece of gear and start again is probably a pretty big ask.

“What is an Eftpos terminal anyway? It’s a data processing tool, that’s all it is. It collects bits of data and encrypts it and pumps it up the line. So you’ve got a modem and some sophisticated thinking processes, a CPU and a display and a few other bits and pieces,” says Bailey.

“What you put into it as an application is totally open-ended. If Joe Blow at the corner dairy wants to put a new loyalty scheme in, he might have a bit more difficulty than Mobil Oil. We’ve got an answer where you don’t have to upgrade the terminal at all, which might not suit Mobil, but there’s different types of marketplace segments, and we’d suggest that upgrading the terminal to provide loyalty should not be a big issue.”

Cooper-Smith argues that upgrading and certifying existing terminals is going to cost money, with some schemes adding “as much another 12 % on the value of the terminal, by the time you’ve got it all capable. I can tell you now it’s the merchant who’s going to have to front that bill eventually — the person who owns the terminal. The manufacturers aren’t going to wear it. Nobody out there is seeing an increase in terminal demand as the result of smartcard technology.

“The other thing I’m really concerned about is how on earth are the customers going to be able to make decisions at the point of sale with the existing terminals, with all those multi-function buttons? I saw a VeriFone terminal recently that had four different stored value card applications running on it. You had to push a different button for each implementation, and each button also had a different function in native mode.

“Hypercom’s view is that the future lies in creating a much more interactive terminal for the customer. We’ll be releasing touch-screen terminals, where the screen changes to suit the need at the time. You have to realise that the consumer is going to be making a lot more decisions at the point of sale in the next few years. That’s going to be a very powerful thing for the merchants.”

So why not ditch it all and go with a Network Computer, or a PDA like the Newton Message-Pad 2000 — both of which have been deployed in smartcard-based electronic commerce trials elsewhere in the world? Then, perhaps the POS network could catch up with the emerging worlds of CORBA and DCOM.

“Exactly,” says Cooper-Smith. “If that’s where the whole world is going in the information technology business, why on earth are the point-of-sale people still trying to introduce a technology like smartcards on such an archiac thing as the SAM? Now we’ve got Multos, and we’ve got four different purses out there that all need separate SAMs — and then it’s all offline. What about the fraud? We know very well that most fraud has happened in an offline environment.

“We don’t for a minute expect that this client-server platform can be deployed immediately, because these cards are pretty chatty backwards and forwards to the SAM. But client-server is definitely a key part of the future and we are going to work consistently forward on the basis that it becomes a client-server environment eventually, and when you want to be able to add another loyalty programme capability to a terminal, it’s done on a server — and you don’t have to go out there and add another SAM.”

Join the newsletter!

Error: Please check your email address.
Show Comments

Market Place

[]