¿ìè¶ÌÊÓƵ

Electronics and computing are big in Europe. By the end of the century,
they will form the largest single market sector in Europe. The world market
for electronics and information technology has doubled in the last four
years alone to 700 billion Ecus ( £500 billion). But in the same period,
the European Community’s trade deficit in the sector doubled to 31 billion
Ecus. The European Commission is worried, not only because this drags the
trade balance down, but also because IT underpins many other markets, including
telecommunications, the car and service industries.

The Commission’s antidote to growing US and Japanese market dominance
is the European Strategic Programme for Research in Information Technology.
ESPRIT is intended to increase Europe’s research capability, giving industry
a better chance to create new products and compete. This approach has held
fast since the programme’s birth in 1984, but there are fundamental changes
afoot. This month sees the start of negotiations for approval of the third
tranche, or phase, of projects. There are signs that ESPRIT officials are
replacing research with development, and are trading competition with the
US and Japan for collaboration.

ESPRIT has three objectives: to provide Europe with basic technologies
which will make it competitive in electronics and IT; to promote cooperation
between companies; and to involve Europe in drawing up worldwide technical
standards. The programme’s rules state that projects must include industrial
companies from at least two European Community countries, though many projects
also involve universities.

The projects supported by ESPRIT rely on ideas from industry and academia.
If their proposals are approved, the projects are awarded half the costs
of ‘pre-competitive’ research for up to five years. Commission officials
divide R&D into three categories: basic research, pre-competitive research
and near-market development.

Basic research covers topics such as superconductivity and molecular
electronics, where even the basic physics and chemistry are not fully understood.
This aims simply to increase understanding. Near-market development is roughly
equivalent to product development. It takes an idea that clearly works and
turns it into a commercial product. Pre-competitive research is the bit
in the middle: it explores what might be done practically, and how, with
the results of basic research, but does not have a competitive edge.

ESPRIT’s survival to a third round shows that the Commission believes
it is a success and gives value for money. In its first two rounds, ESPRIT
has ploughed 4.7 billion Ecus into pre-competitive research (though some
of this money will not actually be spent by project participants until 1994).
This amounts to about £600 a minute.

The programme has certainly been successful in encouraging collaboration
between competing companies, between companies and academia, and between
suppliers and customers. Frits van Hout, the international projects manager
at ASM Lithography of Veldhoven in the Netherlands, says that this has been
the most valuable aspect of his company’s involvement in ESPRIT.

ASM Lithography is one of the partners in a project to design wafer
steppers – machines which make the tiny circuits inside electronic chips.
Some of the company’s European customers and suppliers were also partners.
‘Apart from the financial benefit – which is not major – there is the close
cooperation and the early feedback from the users who got the machine earlier
than they would have done in a commercial situation,’ says van Hout.

ESPRIT money has encouraged firms to do research which they would otherwise
not have been able to afford. Carlo Cibei works for the Italian robotics
company Elsag Bailey, which has collaborated with other companies on a computer
vision project. ‘I believe none of the partners could have done the work
alone. Nobody could have afforded it,’ he says. One Commission official
went so far as to say: ‘If it weren’t for ESPRIT, we would have no IT or
software industry in Greece.’

Brian Oakley, who was the head of the Alvey programme – Britain’s former
national IT research initiative – and is now a director of the software
house Logica, says the greater cooperation spawned by ESPRIT has long-term
benefits too. Today’s young researchers will be the heads of industry in
10 years’ time, and because of ESPRIT will know their counterparts across
Europe.

Beyond cooperation, the programme is also widely acknowledged to have
been successful in the tricky business of setting worldwide technical standards.
But neither of these make a dent in the trade balance; to do this needs
new products.

In 1984, the Commission said that new products would emerge from increased
pre-competitive research. But in 1989 ESPRIT’s focus shifted to include
a Basic Research Initiative. Research under this banner is intended to supply
preliminary findings which can then be developed in other ESPRIT projects.
Work has been done in areas such as superconductivity (resistance-free electronics),
nanotechnology (motors and mechanical devices smaller than the width of
a human hair) and new materials for use in optical communications.

Oakley thinks the initiative has proved itself: ‘It is a real success.
Without it we wouldn’t have the basis for the next generation of projects.’
He acknowledges that the Basic Research Initiative signals a substantial
shift in ESPRIT: ‘It’s a real change.’ Not only are the projects different
in content from ordinary ESPRIT projects, but there need not be any industrial
participants.

More controversial than basic research is ESPRIT’s subtle shift towards
funding near-market research. This is the first concrete sign of a victory
won last year by the commissioner for IT, Filippo Pandolfi, who favours
market intervention, over Leon Brittan, the commissioner for competition
who advocates a ‘hands off’ approach (This Week, 6 April 1991). Some of
the projects in ESPRIT’s second round have already yielded saleable products.
And in the third round only projects which are ‘more tightly focussed’ will
receive funding, says ESPRIT’s director Jean-Marie Cadiou. Other programme
officials say that R&D alone cannot produce successful companies.

Existing participants have also called for ESPRIT to move towards funding
near-market research. Cynics say that these firms are asking for more funding
simply to see how much they can get for free. But the companies give good
reasons for asking for more. Some successful research projects have remained
just that – companies participating in these projects say they could not
afford to go it alone and develop products.

Cibei of Elsag Bailey says that a lack of money made it difficult to
move the results of the computer vision research into production. ‘We could
have exploited this project more if the EC had helped,’ he said.

Oliver Gajek of Siemens Nixdorf Informations Systems in Munich, who
is working on a computer translation project, believes that ‘it would be
advantageous if the Community could fund overtly commercial technologies.
The Community could be very helpful with medium term financing.’ He said
there could be a case for the Community setting up an innovation bank, especially
since venture capital is virtually unknown in Germany: ‘We need to protect
the investment the Community has made (through ESPRIT). It would be different
if Europe’s IT industry had lots of money.’

One researcher privately sounds a note of caution about near-market
research. ‘I used to work on an ESPRIT project that was devoted to developing
a product,’ he says. ‘Each partner wanted to have a product for their particular
application.’ This selfish approach made it nearly impossible to make decisions
and his company finally pulled out.

However, other projects have fared better. ASM Lithography has already
sold several of its wafer steppers and has begun another ESPRIT project
to develop the next model. Officials insist that this is not product development
because the wafer steppers are research tools handling only a few wafers
per hour, and not commercial scale machines. But this argument could be
used to justify funding the product development of many small-scale devices.

Oakley thinks near-market research is fine. ‘Why shouldn’t (ESPRIT)
support marketing at the same time as R&D? If we’re prepared to distort
the marketplace by R&D, why not by marketing too?’ Near-market research
is, however, expensive, he says. ‘The scale of funding is vastly larger
than the ESPRIT budget.’ He quotes a rule of thumb that research costs one
unit, development 10 units and putting a product on the market 100 units,
‘and we’re talking about the 100 end.’

Oakley berates scientists for thinking that marketing is nothing to
do with them, and suggests that ESPRIT projects should include market analysis,
since a brilliant product in the wrong market will fail. Mike Hobday of
the Science Policy Research Unit at the University of Sussex disagrees.
He says market analysis is nothing to do with near-market research per se
– it’s just proper preparation for research.

He is also disturbed by the notion of funding product development. Even
if the practice gets past Leon Brittan within the Community: ‘it could lead
to a subsidy race and could be contrary to GATT (the General Agreement on
Tariffs and Trade),’ he says.

Last March, after the British computer company ICL had been taken over
by the Japanese firm Fujitsu, the Commission ruled that ‘any company established
in the European Community’, could participate in ESPRIT projects. Since
then, programme officials have actively sought collaboration with countries
outside Europe. The Commission is negotiating with the European Free Trade
Association, the US, Japan, Canada and Australia on joint research into
intelligent (computer controlled) manufacturing systems, and a study into
the feasibility of collaborating is due to begin soon.

The successful tie-up between IBM and Siemens – outside of ESPRIT –
shows that international collaboration can be very successful. Earlier this
month, the companies manufactured the first production prototype of a huge
64 megabit memory chip, ahead of the Japanese.

Oakley believes that collaboration with non-European nations can often
be a good idea. ‘If a business is global, there’s no point staying within
Europe.’ Hobday also speaks out against the notion of ‘fortress Europe’.
‘Maybe opening up ESPRIT (to the US and Japan) will give us the opportunity
to interact and learn the tricks of the trade,’ he says.

But can European firms come out on top in collaborations with, for instance,
Japanese companies? Japan has, after all, a formidable record in exploiting
pre-competitive research. One official within ESPRIT says: ‘Companies should
be reassured that they will get as much as they give from Japan.’ He argues
that there is a ‘natural limit to collaboration’. Companies that join forces
in ESPRIT today are cut-throat competitors, and guard trade secrets jealously.
So the joint research ceases naturally when a product gets near the market,
and it doesn’t really matter who collaborates.

ESPRIT has undoubtedly been successful in some areas. Cadiou says in
the organisation’s 1991 annual report that the programme has generated nearly
500 significant results, culminating in commercial products, better manufacturing
processes and improved standards. ESPRIT projects involving the transputer,
the British-designed parallel computing chip, have alone spawned around
500 products.

However, on ESPRIT’s goal of increasing Europe’s ability to compete,
Cadiou seems doubtful. ‘The continuing survival of an independent European
capability in (IT) is far from certain . . . The financial amounts (from
the Commission) currently involved are too small to compensate for the enormous
advantages which IT companies enjoy in other parts of the world.’

Oakley, too, sees international competition as ESPRIT’s Achilles heel.
‘We’re not capturing world markets with our IT, so ESPRIT is failing in
one way,’ he says. Hobday believes the notion that ESPRIT could reduce Europe’s
IT trade deficit is unrealistic. It could only make an impact in the very
long term. ‘What affects market competitiveness is good management and forward
investment,’ he says.

One ESPRIT official argues that eight years is too short a time in which
to measure success. ‘It took 20 years for Airbus to become profitable, and
20 years for Ariane (the European space rocket) to be successful,’ he says.
The comparison between companies that make aerospace products, with lifetimes
of 30 years or more, and IT products, which tend to be overtaken in less
than three years, may be overoptimistic. Nevertheless, his method for gauging
the success of ESPRIT up to now is brutally realistic: ‘Europe would have
been in a much direr situation if ESPRIT hadn’t been created.’