A strong hand for the chip set

Are smart cards the key `enabling technology', as some experts claim? Mark Vernon investigates

Smart card is technology "on the verge". At least, that was the phrase being used all over the Smart Card 98 exhibition and conference last week. The question is, on the verge of what?

For years, credit card-sized pieces of plastic embedded with silicon chips have been "on the verge" of breaking into our lives. And this year's show, bigger than ever and occupying all four floors of London's Olympia 2, came complete with the latest generation of statistics to support the perpetual claim: 98 million cards in circulation by 2002 in the UK, according to Datamonitor.

But to the seasoned observer of this annual event, witness to the frenetic activity of technology players incessantly forming new alliances, producing new standards and running new trials, all in the hope that at last it is about to happen, you have to ask whether the industry is on the verge of a nervous breakdown.

The reason why not is that the fundamentals are good, if short-term turbulence is high. Smart cards are a certain bet; even though no one is yet quite sure which will be the winning horse, that is a minor detail in a field in which pretty much everyone is backing pretty much all of the consortia.

The Information Age offers the firm foundation upon which the industry takes its brief spells of rest. No one can doubt the convergence of computers, telecommunications, entertainment, public and retail services in the networks being wrapped around the world, which means that customers increasingly expect mobility, accessibility and personalisation in the services they use.

"Any device is `my device', wherever I am in the world," as Jon Collinson, strategic and business development manager for BT Consumer Mobility, put it in a keynote presentation. The only practical way to deliver this is to be able to plug a smart card into whatever terminal happens to present itself, from which the necessary data or application can be read, so that the customer can get, do or find what they want, as they want, when they want it. To put it in the jargon, smart cards are the key enabling technology.

If the theory seems a little idealistic, the reality is that developmental activity is consistently high. National governments are running pilots to move public services online, which requires smart cards to identify citizens. The set-top boxes for digital TV will contain two smart card readers, each to assist in the payment process. Most banks will have slipped chips into debit and credit cards by the end of this year, ready for the spread of a commercially viable retail infrastructure. Mobile phone companies, already big smart-card users carrying the GSM standards, are busy forming partnerships with electronic cash systems to lever their investment further. Even those old enemies, MasterCard and Visa, have virtually become mere sparring partners; recent noises from the boards of both suggest it is time to make amends and get down to serious business.

This last development is an important one, and it leads analysts such as Mark Stevenson from Ovum to predict that we will witness a smart card "small bang" by the end of this year. What has happened, he believes, is that the proprietary days have come to an end, when developers tried to keep a hold not only on service applications but also on the operating systems which ran the cards themselves.

"Suppliers are trading off positions of power as providers of everything, to open up the market, though in a way which they hope will allow them to become the biggest," he says.

Two standards in particular have come to the fore: Maosco's Multos, an operating system designed in particular to enable smart cards to run multiple applications, widely recognised as a significant driver, and the JavaPOS initiative, aimed at making Java the language of choice. A number of important hurdles are bridged with both solutions. For one, they enable applications to be written in higher programming languages, Java and C, well understood across the industry. Very few people can successfully program the core layers of smart card technology, largely because the code has to be so concise.

These initiatives also allow the industry to adopt a familiar model for the evolution of the technology, based upon the development of the PC. But there are differentiators, too. Perhaps the most illuminating concerns the hype that surrounds any development that claims the Java badge. Sun Microsystems' Java mantra, "Write once, run anywhere," is in reality tricky to fulfil, even with the power of the PC to back it up.

The trouble with smart cards is that the term "smart" is relative to its monosyllabic cousin, the magnetic stripe, not the processors for which Java was designed. When squeezed on to smart cards, the language is inevitably compromised. As Ovum's Stevenson says, "It is not a standard platform, and is not inherently secure in itself." Though he is keen to stress that it can be made so on both accounts, it just takes some effort.

The story of IT is full of high expectations, some of which lead to glory. Smart cards are confident in the former and expect the latter, though with a certain seriousness to boot.

In fact, there was only one real gimmick at Smart Card 98, a rare thing for a computer show: a gold-laminated smart card. How one must dress to impress in the information age.