Trying to come up with tomorrow today

Getting paid for imagining the future is the kind of job description that can tempt a computer scientist to forgo even share …

Getting paid for imagining the future is the kind of job description that can tempt a computer scientist to forgo even share options.

Creating good software and hardware can be deeply satisfying, but there's a special pleasure in actually pushing the conceptual boundaries - thinking of what might be, then figuring out how to get there - using the power of the microchip.

That's the job of researchers like Mr Philippe Janson, manager of research at IBM's Zurich Research Laboratory, one of eight the company operates worldwide. In Dublin recently as part of a conference on the future of technology, Mr Janson has been an IBM man since he left the Massachusetts Institute of Technology (MIT) in the 1970s with a computer science degree.

Then, as now, attending MIT meant submerging oneself in a blast of technological innovation. Mr Janson says: "As they say there, it's like trying to drink from a fire hose." But that's useful preparation for working in a research laboratory. Zurich boasts up to 300 researchers at any one time, all trying to come up with tomorrow today.

READ MORE

Where IBM laboratories in past decades had more of a pure research tangent, a shift in the 1990s towards greater customer focus means there is greater likelihood that current projects will be in a box on the shelf at some point in the future.

To prove this point, Mr Janson outlined an array of current IBM research directed at electronic commerce. "E-business means technology that makes it possible for a business to interact electronically with its suppliers and customers," he said. Some of the technologies are easy enough to understand, such as a writing pad co-developed by Cross, the pen company. A normal pad of paper sitting upon a special thin computer appliance reads the impressions of a pen and translates written letters and numbers into text in a computer file, and creates line drawings from any illustrations.

A number of Web-based technologies also seem to be very practical. Hotvideo, for example, allows video images on a computer screen to be interactive, just as static images on a Web page can be. Clicking on a video image could lead to text files, another related video clip, or a database.

PanoramIX is a three-dimensional application that allows viewers to explore an environment - a floor of a department store, for example - by creating a seamless virtual location out of multiple photographs.

But other technologies verge on science fiction. For example, IBM discovered earlier this year that the human body could act as a medium for transmitting radio signals, which launched the concept of the "personal area network".

Mr Janson explained that a radio transmitter worn by two people could enable one person to send an electronic business card, or any other information, into the transmitter carried by the other simply by shaking hands.

Flash (an abbreviation for Fast Look-up Algorithm for Structural Homology) is a special mathematical formula, or algorithm, which enables complex structures like DNA or fingerprints to be sorted through and matched. As long as someone can find a system for describing an object accurately, any structure can be analysed, said Mr Janson. The technique is already proving itself as a forensic tool because it can quickly analyse thousands of fingerprints to look for matches. It can also be used for molecular matching or for searching through large amounts of text.

Another application, called Teiresias, can examine massive "bitstreams", or the streams of characters generated by a computer program as it executes and finds the repetitive patterns that define the normal functioning of a computer program. In a test project, software was designed which learned those normal patterns, then was put to work watching an operating computer for any abnormalities.

Abnormalities would be likely evidence of an attack by computer hackers attempting to break into a system.

"If you teach the [software program], these are the good patterns and let me know any that aren't, the system can raise the alarm in an attack," said Mr Janson.

Smart cards, or plastic cards similar to credit cards which contain computer chips, will also be used in innovative ways in the future, he predicts. "Today's smart cards typically carry a single application," said Mr Janson. "Tomorrow's smart cards are going to be fully programmable little animals."

As an example, he held up his IBM identity badge. "Wouldn't it be nice if that same badge could sign me on to my computer?" he asked. Then, imagine that Visa or Mastercard could be carried in the card, a card's owner could purchase a plane ticket with the card on the Internet and have the ticket transmitted onto the card. The same card could be used to book a hotel, then used to open the hotel room without having to formally check in.

After a business trip, the records of all purchases made with the card could be downloaded onto a computer to generate an expense report.

But Mr Janson, who is particularly interested in computer and network security, isn't a blind devotee of technology for technology's sake. He says he is wary about how technology can be used to exploit rather than enrich.

"I think every technology - every one of them - brings its own risks," he says. "It's up to us technology guys to contain its use."

Karlin Lillington

Karlin Lillington

Karlin Lillington, a contributor to The Irish Times, writes about technology