THE NEXT STEPS: after the UK Workshop on Background.

THE NEXT STEPS: after the UK Workshop on
Grand Challenges, Edinburgh, 24-26 November 2002
The purpose of the Workshop on Grand Challenges for Computing Research is to
discuss possibilities and opportunities for the advancement of Computing Science,
particularly in UK. Its method is to solicit submissions from the UK computing
research community, identifying ambitious, long-term research initiatives that might
benefit from some degree of national and international co-ordination. In other
branches of Science, such projects are known as Grand Challenges.
Experience shows that greatest benefit is obtained when a challenge is driven by
scientific curiosity or technological ambition of the scientists themselves, and when it
commands the support of many other scientists, even outside the research groups who
are working on it. A Grand Challenge is therefore complementary to other kinds of
special research initiative, which are often devoted to activity in some particular
research area that has the potential to contribute to the goals of a particular nation.
Often, the goals are non-scientific, and include elements of commercial, military,
political or social benefit. A Grand Challenge also differs from this other kind of
challenge in its longer time-scales, in its prior formulation of a clear and testable
criterion of success or failure, and in its setting of a target end date for completion. If
financial support is committed, the scientists promise that it will terminate on or
before that date, whether the Challenge succeeds or fails.
The report of the Workshop will be submitted to its sponsors, the UK Computing
Research Committee in two stages. The first is a brief account of conduct of the
workshop itself to be submitted to the next following UKCRC meeting. The second
stage will occur after a six-month discussion period, conducted by contributions to a
shared website from the UK computing research community. We hope that the final
report will concentrate on not more than a dozen Grand Challenge proposals, each of
which has a broad constituency of support. They will be drafted in a somewhat
uniform format that will assist their summarisation, comparison and evaluation
according to a variety of criteria. The principal criteria will be closely linked to the
maturity and feasibility of the challenge:
What are the ultimate ‘deliverables’ in ten or fifteen years? How will we know if the
project has succeeded or failed? How does the project split into work packages with
intermediate objectively determined targets? What research methods will be
relevant? Is there a body of researchers trained in those methods? Who are the
principal international figures and teams in the area? Which work packages will be
the first to start? Is the state of the art yet ready for the launch of a major planned
Discussion Period.
During the discussion period, the UK computing research community will be invited
to make contributions of three kinds:
(1) Suggestions and criticism of the concept of a Grand Challenge, and how it has
been interpreted in the output of the Workshop. (2) Improvement to an individual
grand challenge, to take advantage of research skills and methods offered by the
contributor. (3)New grand challenges, drafted in accordance with the uniform
To promote such aggregation of support for unified proposals, the Programme
Committee has made an initial classification of challenges into areas of affinity.
Later, we may conduct a questionnaire among contributors to assess the strength of
feeling in favour of each potential merger. The main questions would be:
Is there a Challenge to which you would like to devote your own personal
research for the next ten years or more?
Which Challenges outside your own specialist area of research do you think
would contribute most to the advancement of Computer Science, and the contribution
to it made by UK?
We hope that contributors will not under-estimate the effort and commitment required
to formulate and plan a Grand Challenge. It is like writing a research proposal to be
judged by the entire scientific community, for a project that will last ten years, that is
to be conducted by teams at ten sites distributed round the world, and which may
consume a thousand man-years of scarce skilled research effort. And there is no prior
commitment that anyone will take any notice of the proposal. Many serious scientists
will prefer to have nothing to do with the planning, administration, and routine
groundwork associated with a Grand Challenge. It would be quite reasonable for
many of the authors of existing submissions to decide that they belong to this
majority of non-participants. On the other hand, for those participants that remain,
we see tremendous opportunities for collaborative scientific achievement; some of the
early steps towards the achievement are outlined in what follows.
Substantive report to UKCRC.
At the end of the six-month discussion period, the Programme Committee will
attempt to summarise the Grand Challenges that evolve from the six-month
discussion period. They will attempt to assess the likely timescale for realisation of
each Challenge, and recommend which ones should be pursued more urgently than
the others.
The UKCRC then has an opportunity to make their own assessment. Their criteria
may include
Potential contribution to the advancement of science.
Feasibility and timeliness as a basis for action on an international scale.
Potential for UK contribution and even leadership in the field.
For any challenge that seems appropriate for immediate action, the UKCRC may
sponsor a Workshop for potential UK contributors to that Challenge, selected from
those who have contributed to the discussion. It is important also to attract wider
participation by scientific leaders from round the world. The purpose of each
Workshop would be to formulate and progress concrete plans towards the fulfilment
of the particular Challenge. UKCRC has already sponsored a Workshop of this kind
on the topic of algorithms and complexity.
It should be emphasised that there will be no implied criticism of any Grand
Challenge that is not pursued immediately. We are planning for the long term, and
there will be other opportunities for revival of postponed challenges. We hope that
the UKCRC and EPSRC will find some mechanism by which new Grand Challenges
may emerge, ensuring that challenges in progress will be monitored, and old ones will
be discarded on their success or failure.
Implications for fund awarding agencies – the EPSRC.
The deliberations and the conclusions of the UKCRC will be forwarded to the
relevant Strategic Advisory Team (SAT) of the EPSRC. The adoption of a Grand
Challenge may contribute to the goals of the EPSRC in the following ways:
1. It will announce the intention of EPSRC that the UK plans to be a leading player
in an international scientific venture. If the venture is successful, it will boost the
international reputation of UK Computing Research.
2. It will contribute to the recently adopted EPSRC policy to support more ambitious
research on longer timescales. This can be achieved by a succession of quite small
research projects, selected by the existing panel mechanisms on the basis of their
cost-effective contribution to a common long-term goal.
3. It will improve the quality and reliability of the selection procedures. Any proposal
that claims to contribute to an adopted challenge can be judged solely on the promise
of its technical and scientific methods, without the need for further justification of the
overall goal.
4. It may also assist in selection of referees, since they can be chosen from those who
are known to support the general goals of the challenge, and be competent to judge on
the merits of the means.
In these ways, it should be possible to promote a small number of Grand Challenges,
without any change to the current selection procedures for responsive mode research