BELIEF project outcome...Brainstorming the future of e-Infrastructures

By Stephen Benians, BELIEF –I and II project coordinator, Metaware SPA,  Italy

Two opinion pieces inspired by the many ideas that emerged from a recent experts’ focus group in Brussels: the BELIEF project’s 5th brainstorming event . Rather than concentrate on creating impact metrics for use in the future, we need a census that maps out evolving e-Infrastructures usage and applications, so policy can nurture evolving visions, not prescribe a given future. This helps to sell the value of advanced e-Infrastructures as an open innovation platform that drives future development, an essential point to drive home when trying to secure new FP8 budgets.
e-Infrastructures census: nurturing the future
The US is embracing its ‘census 2010’ with gusto. Not just vital for measuring social and economic shifts, the census is proving a helpful way to sustain community level employment. But just as the US started rolling out its 2010 edition, so a new type of census has got people talking in the EU: the e-Infrastructure census. An experts working group concluded that a census is key to justifying, sustaining and maximizing the value of e-Infrastructure investments. The census logic goes like this: We all use public services. These services need to be planned, to keep pace with the fast-changing patterns of modern life. But it is impossible to offer services to ‘invisible’ people – we need accurate information on the numbers of people, where they live and work, and what their needs are. Since the same questions are asked each time, dynamic trends emerge to which policy can react. The same logic, say the experts, must be applied to ‘set the record straight’ on future public policy and usage of e-Infrastructures.
The vision of e-Infrastructures is to boost collaboration among research communities and empower them, through ubiquitous, trusted and easy-to-access services for data, computation, and communication. Applications leveraging e-Infrastructures in domains such as e-Science, e-Health, or e-Education, have informed of great opportunities and novel ways to produce and share new knowledge, establish cross-disciplinary connections and diffuse research outcomes into the society.
However, despite their considerable technological advancement and proliferation, e-Infrastructures have not yet reached their envisioned impact. This weakness is mainly due to external socio-economic factors.
To reach the ultimate goal of productive Global Virtual Research Communities, it is therefore important that readiness, penetration, quantitative and qualitative usage patterns of e-Infrastructures be regularly monitored by governmental bodies and other related initiatives. The census would be the tool for taking the pulse of e-Infrastructures usages and development.
This may all sound rather agreeable but before such a valuable tool can yield fruit, the census’ design would require careful consideration. Since virtual research communities rely to different degrees on the several network, grid or data layers serving them, it is hard to attribute ‘impact’ to ‘e-Infrastructure’ investment per se. One of the major findings of the expert group was therefore that each layer of the e-Infrastructures must be evaluated with its specific metrics.
The network layer already benefits from the REDI indicators. However its social impact, although undeniable, is best measured in negative terms: what if it were taken away? A similar analogy could be made to ‘roads’ as a public service. In developed economies, everything from hedge-funds to hospitals would collapse without them. However, fund manager and surgeons would not normally attribute their success on any given day to the quality of the roads. In the same way, the social impact and the contribution of e-Infrastructures to EU strategic visions cannot be attributable to just the networks. Instead, the closer the e-Infrastructures ‘layer’ is to the user, the more attributable the e-Infrastructure is to the final impact. If we want to ensure each layers’ critical role is recognized (and therefore funded!), then linking metrics are needed, to show the intensity of causal links and dependencies between each layer, from the network through to an end user’s discovery (‘from backbone to Bingo!’) .
When it comes to grid, there is already an array of performance metrics, though any indicator adopted would have to depend on which research community, or individual researcher, is using the grid. The measurement of data exchange however (how open, how reliable is it?) as well indicators of user satisfaction, is a very different, evolving issue and no solid metrics have yet been identified. This is a major issue since future investments depend on proving the utility and value of these infrastructures. And this is where the census comes in. The experts working group laid out a possible terms of reference of a tender for a pilot census, to be then mainstreamed through an EU agency such as Eurobarometer, once when tried and tested.
The terms of reference set out how to define the user and their needs; and recommend which questions will yield answers that help policy makers steer strategic paths and governance structures of e-Infrastructures into the future. The results will be critical for attracting investment. But not only: they census results will be key to highlighting business opportunities and cross border partnerships that are evolving. And as well as consolidating the role of e-Infrastructures in competitiveness, the e-Infrastructures census, like in the US, can contribute to economic revival in the EU.
Metrics must give way to Visions
Metrics are a dangerous way to sell the idea of investing in e-Infrastructures. These technologies have created most value by fulfilling new, unanticipated user needs. This is not specific to e-Infrastructures of course. Most of the value generated by Nasa’s space programme was from investing in experimental technologies that only later fulfilled previously unimaginable needs on a large scale. The value of e-Infrastructures specifically lies in the freedom to innovate, in helping to realise new visions that evolved since the technology’s inception.
Metrics can be useful - for measuring Incremental innovation (ie, cheaper, thinner, faster more features) and ‘total factor productivity’. But social and scientific dividends (or ‘profits’) from an e-Infrastructures innovation can be a long time coming and depend upon many factors outside the innovator’s control, or imagination. Hence the need for a periodic census sustained over time, to pick up on valuable impacts and their footprints back to a particular innovation. But if the real value of e-Infrastructures lies in therefore their ‘innovation potential’ how can we measure this, the “freedom to screw up”? A toolkit of indicators to determine this must steer clear of the binary factors of ‘implementation’ (inputs like finance, technology) and impact (like scientific and societal benefits) of e-Infrastructures. Bearing in mind that the most disruptive innovations over the last centuries have been those that change how we are run (institutional revolution) as well as the new activities that such a new order allows us to engage in, then any new toolkit of indicators should measure both factors of ‘governance’ and ‘innovation activity’ together. This is where indicators of user satisfaction will be key; it is after all the researchers, the end users themselves who need this freedom. However, again, this will only be of value if measured over time, in a census. So to win FP8 funds for e-Infrastructures on the back of a ‘results presentation’ to member states would not only be misleading, but it be less impressive and ultimately less yielding, than a bolder attempt to demonstrate their innovation potential over time.
See www.beliefproject.org for more details on the brainstorming that inspired these ideas, and for the final report that will set out recommendations emerging from the event.