Mining for knowledge

When I took on the task of Chairman of what was then the newly formed Engineering and Physical Sciences Research Council (EPSRC) five years ago, it was a time of considerable anxiety in the academic community because the term ‘relevance’ had been introduced into the rationale for state-funded scientific research. There were fears that this […]

When I took on the task of Chairman of what was then the newly formed Engineering and Physical Sciences Research Council (EPSRC) five years ago, it was a time of considerable anxiety in the academic community because the term ‘relevance’ had been introduced into the rationale for state-funded scientific research.

There were fears that this would lead to short-termism and that the national science programme would be allowed to deteriorate into a development activity for industry, and I guess my presence as an industrial person didn’t help much.

There is no doubt that we need a clear rationale for government investment in academic research and that there should be some clear expectations with regard to the output. However, there are a number of people who would try, I think mistakenly, to justify this investment primarily by the inventions which arise as a consequence.

Universities are also encouraged to try to optimise their financial return on inventions and intellectual property and in doing so run a danger of upsetting, or at least constraining, their relationship with their customers in industry. This is despite the fact that, on average, royalties produce only 1% of the total revenues of the university. If you look at the benefit universities can gain when they have a good relationship with industry, it is worth far more than the 0.3% of their revenues.

So I would argue that the narrow focus on inventions, intellectual property and technology transfer is dangerous. It is based on a misunderstanding of why government funds academic research and it can, when not placed in context, be damaging to the science base.

To my mind a more realistic argument for government spending on academic research would be expressed by the following: government invests in scientific research primarily to generate that knowledge and expertise which will enable the nation to deal with the next generation of change.

If we accept that this is the purpose and that the flow of knowledge and expertise is the key objective, then we can begin to define on a more logical basis the respective roles for government, the research councils, academia and industry in the overall process.

The national research cycle is a linked process involving government, its agencies, such as the research councils, the providers of research in academia and elsewhere and the users of the research in industry and elsewhere.

Each of the contributors has a distinct role. Government sets the broad policy for scientific and engineering research via the Office of Science and Technology (OST) and through White Papers. The Research Councils implement the policy of Government within the funding provided. The EPSRC generates relevant new knowledge and expertise within its broad area of responsibility and to ensure that there is a healthy outflow into the community at large.

The universities’ role as research providers is knowledge generation both at the programme and project level.

Industry is the user of research. It is the recipient of knowledge in that it draws upon the graduate and postgraduate products of the system and upon the relevant knowledge and expertise which is generated.

However, industry can improve the knowledge flow with mutual benefit by providing feedback, by interacting directly with researchers in the academic sector, and by funding research and knowledge-transfer activities.

The EPSRC has looked closely at its relationship with its principal customer, the OST, and with the providers and users of knowledge. It has established a new structure and policy mechanism. As a result, it now oversees the shaping of a landscape of broadly defined research programme areas and seeks to establish a balance of funding between them.

This process is repeated annually to look at the balances between these areas and to decide the prioritisation between the broad programme areas on a year-by-year basis.

The EPSRC then calls for proposals to populate this research landscape with projects which originate from the academic sector. It processes the proposals from the academic sector and uses peer review to select and prioritise the proposals. It then funds the activities and organises the necessary reviews and audits to ensure that the research is performed satisfactorily. In the past there was far too little attention on evaluation of the research activity and we are trying to alter this balance.

But what about relevance, and how is this to be judged at the research proposal level? To answer this question one needs to understand some basic truths about research projects.

First, a research project’s genuine value cannot be usefully determined at the time of its completion. It may be months, years or even decades before the value of a piece of research is apparent and even then its use may be dispersed among many applications. The case of ‘eureka!’ in the bath and ten years later 50,000 blue widgets is very rare.

So what can you determine in the time frame of the project? In the short term you can assess the following: whether the researchers are professionally competent; whether the research has been performed to the highest possible standards; whether mechanisms for outflow exist; and whether the work has been performed in an area that can be recognised as broadly relevant.

Essentially you are setting people to mine knowledge in a broadly relevant area. For my money, that is where relevance comes in. If a large proportion of the knowledge generation process comprises a portfolio of activities with competent people mining knowledge in a broad area of relevance and with mechanisms in place for outflow, then I believe this is the best that can be done.

I also believe that if it is done well, with appropriate metrics, then it will give you a significant advantage over a random system, where everybody does what they like.

Alan Rudge is the chairman of the Engineering and Physical Sciences Research Council. He finishes his five-year term of office at the end of March.