Hugh Aldersey-Williams on a radical approach to innovation, born in a Soviet patent office
Try this puzzle. An empty bottle is stood upside down on a piece of paper which lies on a table. How do you remove the paper without touching the bottle?

At a party, you might suggest simply whipping out the paper to leave the bottle standing. Researchers at a lavishly funded R&D centre might build a powerful magnet to levitate the bottle.

Both these answers come up if the problem is given to a new software package designed to promote innovation. But so does a better one: use vibration. By banging repeatedly on the table the bottle is intermittently lifted from it, enabling the paper to be pulled gently free.

The TechOptimiser software has been developed by Invention Machine Corporation to facilitate product development at the concept stage. The latest version was launched in this country this month. Aimed at scientists and engineers, the pounds 6,000 package is based on a simple idea.

The story of the company, now based in Cambridge, Massachusetts, began in the Soviet patent office in Baku. It was here during the Stalin era that one Genrich Altshuller first thought to itemise the principles behind the military patents he was processing. After studying thousands of patents, Altshuller came to the conclusion that there were about 40 important principles - such as applying an action intermittently, or applying it briefly but to excess, each of which provides one solution to the bottle problem. The database was later extended to more than 2.5 million patents to create the present software.

This data is organised in such a way as to help today's innovators arrive more quickly at the best solution for their problems. The principles are generic, and the software is already used by organisations as varied as Nasa, Motorola and Oral-B in the United States. British users include Shell and Raychem as well as big names in aerospace and pharmaceuticals.

Rank Xerox expects to load its TechOptimiser discs later this year. The previous version of the software has already shown its worth at Xerox in the US, although it has not swept all before it, finding use only on a handful of projects since 1994, according to Les Wynn, industrial design project leader at Rank Xerox's technical centre in Welwyn Garden City. It is best suited to engineering, where it can provide a breathing space to try something new. "It's a kick-start to large corporations which tend to do what they've always done, especially with the pressure on time to market. But this makes people stop for a second and points out a lot more solutions," Wynn says.

A genuine problem familiar to many is the pizza that arrives soggy after a twenty-minute ride on a moped whose driver models himself more on Evel Knievel than Henry the Navigator. The solution was an airtight box with dimples on the bottom to lift the base of the pizza slightly free of the cardboard. This keeps the pizza warmer by insulating it from the box and cuts down on moisture loss and keeps the cardboard dry.

The medical equipment manufacturer, Gambro, used TechOptimiser to solve problems of bubbles forming in the blood pumped by its dialysis machines. The software identified the problem as one of gases in liquids generally, and so was able to suggest solutions derived from very different fields such a brewing and petrochemicals. "By seeing those examples the company could see that it was facing the same technical problems," says Anders Killander, Invention Machine's vice-president of European operations. "Any engineer only has a knowledge of 5 or 10 per cent of what is known. But they are using a complete listing of what is possible."

The software assists in discarding the more impractical suggestions. But don't company engineers tend to dismiss unfamiliar solutions? "This is what differentiates the truly innovative companies," says Killander. "Innovative companies say, wow, this can make a difference. The non-innovative ones say, oh no, two more possible solutions to consider."

The solution seems obvious enough after the fact. But getting there requires clear thinking. The software forces this by asking questions that lead to a precise definition of the problem. It then enables users to reach practical solutions that might otherwise be either beyond their experience or which they might eliminate because they seem initially outlandish - for example, a technique borrowed from an unallied field.

Using the software also encourages engineers to begin to think more systematically themselves. In an evaluation at the Massachusetts Institute of Technology, experts using the software came up with almost twice as many solutions to a given problem than their colleagues using traditional techniques. What is more, the quality of their solutions was demonstrably superior after a period of familiarisation with the software. Invention Machine claims TechOptimiser leads to more cost-effective R&D, faster payback, and an increased staff knowledge base.

The first software module emphasises the correct description of the problem, beginning with the reason to innovate. Such a thinking process is always important in new product development, but people are easily distracted from the real issues by extraneous pressures, such as the need to get new product concepts into the pipeline for reasons of competition or the advent of a seductive but ultimately inappropriate technology.

The problem may focus upon parts or actions. In the former case, the software prompts the user to consider whether to add, modify or subtract parts. Within these options there is a further level of examination. For example, there is a sequence of rules for trimming parts - because they are expensive, because they are complex or unreliable, because they are remote from the main functioning of the device, and so on.

In the bottle puzzle, the focus of the problem was upon actions, the pressure exerted by the bottle lip on the paper and the friction between the two. The route to the solution here lies in asking whether these actions need to be augmented or diminished - in this case the pressure of the bottle on the paper and/or the friction between the two must be reduced - and then finding effects that do this.

The effects module then allows the user to explore physical effects which might be used to solve the stated problem. In case the name of some obscure scientific effect does not immediately suggest the answer, each is illustrated with examples based on familiar product types. Kodak, for example, makes a motorised camera which automatically sets the focal length and adjusts the angle of the flashgun reflector for that distance. Wishing to make a cheaper version, the company used TechOptimiser to identify the motor - expensive, remote from the camera body, stuffed with moving parts - as the first thing to go. The problem then became how to retain the variable reflector angle. The winning solution was a piezoelectric linear motor - a solid-state arrangement integral to the camera body in which an electric current causes a material to exert pressure that can then be converted to movement.

The irony is that Invention Machine's own innovation could not have been achieved using its software. It has, however, created a company that is growing by more than 250 per cent a yearn