Today, cars are safer, more comfortable and have higher fuel efficiency than even 25 years ago. However, these improvements are less than the sum of the known innovations in material, ergonomic and energy conversion technologies. Europe has been a hot bed of such innovations but engineers have been inefficient at transforming this stream of basic technological innovation into great products that change our lives.

Innovation efficiency is not about generating patents for basic science. Rather it is about incorporating basic science innovations into products. Information technology should be an enabler that improves innovation efficiency. We've had success innovating by embedding information technology into products like mobile phones. Using information technology to translate other innovations into products has not been such a success.

Two barriers share part of the responsibility: Firstly, manufacturers' natural reaction to competition is to produce at a lower price. There are many well-established methods to squeeze costs. Classic information technology solutions like payroll and enterprise accounting are highly effective at reducing overhead costs. Naturally, they have been a first priority for information technology mindshare and investment.

Secondly, the process of designing products is the same now as in the first half of the last century. It then worked well to produce novel products such as radio, domestic appliances and the aeroplane. Today, design engineers draw on screens rather than on drawing boards. Computer Aided Design helps the documentation task more than the design task.

For real improvement in innovation efficiency, the design process must change. Electronics is the one engineering discipline where that has happened. It is no coincidence that electronics' products, especially computers, have been offering truly astounding rates of improvement. Products are more and more affordable; they are easy, if not easy enough, to use; and each year they are smaller and use less power.

There are only a few hundred thousand electronics designers world-wide, not many more than 30 years ago, but the volume of new products these engineers design has increased by two orders of magnitude on any reasonable measure. In the early 70's, electronics engineering focused on documenting the layout of chips or printed circuit boards and then testing a design prototype. By 1980, automated layout tools generated the physical layout from diagrams describing the logic of design. Engineers simulated designs before prototypes were manufactured. By 1990, designers thought in terms of abstract models describing the functionality needed and let information technology tools synthesise the logic and physical layout from these abstract models. Today, innovation continues with higher levels still of abstract models in design.

Generations of specific design tools for electronic designers supported these revolutionary changes in the design process. They now permit engineers to design products with amazing complexity without needing to visualise and understand the actual layout of circuits on a chip. They are expensive compared with the geometry authoring tools used in mechanical engineering. The innovation efficiency from continuous improvement in electronic designer's productivity justifies the investment.

It should be possible to imagine similar approaches to improve mechanical engineering products from bicycles to printing presses. Current practice has been to use information technology to seek to eliminate or reduce the time on mundane tasks. Within the design authoring task dimensioning is automated. Within project management, automated workflows help manage change control.

The industry hype about Product Lifecycle Management mainly focuses on design infrastructure applications like change control. Ironically, in the electronics design industry, the equivalent methods are not particularly advanced. This has not impeded innovation efficiency. That's not to say Product Lifecycle Management is not highly important and could further improve innovation efficiency. However, it is evidently not the only route forward.

There is a recognition that discrete manufacturers need to compete by offering innovative products rather than simply reducing costs. The well received ProductFirst © initiative by PTC (www.ptc.com) is one software company's contribution to the debate. Many companies are changing their design process to try to re-use parts and other aspects of existing designs. There is renewed interest in mechanical design automation - sometimes referred to as Knowledge Based Engineering. Late last year, Dassault Systèmes, purchased British privately held KTI, a pioneer of mechanical design automation.

This is the visible tip of the iceberg of a new generation of companies offering various mechanical engineering automation tools. Invention Machine helps engineers restate their requirements to create alternative design solutions. Granta Design helps engineers consider innovative materials. Engineous helps optimise trade offs between different simulations for stress and vibration. DriveWorks, IcedCAD, and Rulestream capture the processes used to create complex existing designs and reapplies them to new requirements. A DriveWorks example is a customer specific tanker body built on top of a standard heavy vehicle chassis. An initial design, fully costed and with manufacturing instructions, is generated from parameterised building blocks.

However, these are all unconnected, with each tool only helping specific tasks. An improved design process requires several joined up phases. There are standard openings, where tools help a designer innovate by considering different solution approaches. There is an end game where algorithms generate physical geometry automatically to solve the problem. In the middle, alternative designs are analysed and simulated to identify the match to the requirements and establish optimum trade offs. Today, mechanical design tools offer little support for this middle stage.

There is a great opportunity to improve the efficiency of converting basic technology innovation into products. This can translate into sales, profits, jobs and wealth. Nevertheless, for this to happen, many design offices must change. The focus of the suppliers of engineering software tools will have to shift too. Bigger engineering tool vendors are hyping Product Lifecycle Management to provide efficient design information handling and communication. To win the innovation efficiency race this may not be enough. If our future mechanical engineering products are to take advantage of the complex technologies coming from biology and nanotechnology we will need the same revolutionary changes in the design process seen in electronics.

Mike Evans

First appeared in the FT online Expert's column, August 2003

Other articles from Cambashi:

back to top