An ICRI at the heart of public research

A proposal, and a doubt

When we designed the ICRI, the Invisible R&D Capital Index, our aim was to better understand what makes up an organisation's scientific capital. The skills, the accumulated knowledge, the relationships maintained, but also the opportunities not yet exploited. One question remained: could our method also work in public research?

We suggested to Philippe Drobinski, former director of the Laboratoire de Météorologie Dynamique (LMD), that he test the tool using the data available at the end of his term of office, which were neither confidential data nor personal data. He agreed, without hiding his reservations. In his eyes, the ICRI seemed to have been designed first and foremost for industry.

The LMD, a laboratory that models the sky

The LMD works on the atmosphere and the climate, through observation and modelling. Its teams coordinate, among other things, the development of the LMDZ atmospheric model, the CHIMERE chemistry-transport model and the GEISA spectroscopic database, and contribute to the development of the ORCHIDÉE land surface model. The laboratory is a founding member of the Institut Pierre-Simon Laplace, for which it runs, in particular, the SIRTA observatory, at the heart of the national and European ACTRIS infrastructure. Because of its major contribution to space-based observation, the LMD works closely with CNES, the French space agency. Publications, models, data, doctoral training: its activity is already well documented. We also wanted to look at what escapes these indicators, in particular the know-how and resources that are sometimes lost from view over the years.

With no revenue, what do you measure?

A company can, at least in part, gauge the benefits of its research through its sales. For a public laboratory, results show up in other ways. A model used by another team, expertise called upon by an institution, an industrial collaboration or the career path of a PhD student. Studies of the relationship between public research and industry also show that exchanges take many forms other than patents and the creation of companies. So we could not reuse, as it stood, an assessment grid designed for the private sector.

We revised three points in particular.

First, open distribution. The LMD makes several of its codes freely available. From a strictly commercial point of view, this practice might seem unprofitable. But for research, it allows other teams to take ownership of the tools, improve them and help spread them. This choice had to be recognised in the diagnostic.

Next, reputation. The LMD's visibility stems from its own work, but also from its membership of the Institut Pierre-Simon Laplace. We had to distinguish the laboratory's contribution from that of its environment, without underestimating the importance of the latter.

Finally, missing data. When an answer was not available, we avoided arbitrarily giving it a poor score. Not having an indicator may reveal a need for monitoring, without meaning that the activity concerned is inadequate.

Three months of fine-tuning

The beta test ran over three months, with two complete rounds of exchanges between Philippe Drobinski and Cortex. Philippe Drobinski went through the reports in detail, flagging wording that needed clarifying and interpretations that required more context. The scores did not change, but the way they were read became sharper. These exchanges, together with those held with other academic beta testers, helped our method evolve. Today, carrying out an ICRI generally takes two to three weeks, even for large organisations with a great deal of data. Above all, we seek to produce an analysis suited to the sector, practices and constraints of each organisation.

What the ICRI showed

The LMD obtained an overall score of about 72 out of 100. This is a high score, but what interested us most was the detail.

Its relational capital stood out clearly. With its institutional partnerships, European infrastructures, collaborations in the space sector and the use of its tools by other researchers, the laboratory has a particularly well-developed network.

On the other hand, the diagnostic revealed room for improvement in turning these relationships into new opportunities: industrial partnerships, transfers or diversified funding. The contacts were there, but not every possibility had yet been exploited.

Another finding: part of the scientific assets remained difficult to inventory. Models, data and specialised skills existed, without being brought together in a map that could be used directly to think about their valorisation.

So the laboratory had solid foundations. The challenge was to make better use of them.

The surprise

What surprised Philippe Drobinski most was the precision of the resulting map. The diagnostic connected the teams, their skills, the tools developed and the partners involved. Pieces of information known separately took on a new dimension once brought together.

The report also proposed a progression over time. We recommended starting with skills and knowledge during the first six months, before working on opportunities, reputation and societal impact. The idea was not to change everything at once, but to identify the first useful actions. By the end of the test, his initial reservations had shifted: Philippe Drobinski now saw the ICRI as a cross-cutting tool that could also be of interest to university technology transfer offices and to the Sociétés d'Accélération du Transfert de Technologies (SATT), the French technology transfer companies.

What changed afterwards

One of the most concrete lessons concerned exchanges with the technology transfer units. The diagnostic strengthened the case for an earlier and more regular dialogue. When innovation potential is spotted late, everyone has to piece the information back together in a rush. A map shared upstream can make this work easier. The teams have a clearer idea of what they can present, and the technology transfer offices have the elements they need to examine the possibilities sooner.

What another laboratory can take from it

Public research also has considerable scientific capital. It is simply that its value does not always show itself in the same way as in a company.

This test reminded us of three things. A laboratory with good activity indicators can also have resources that are still little exploited. Valorising a result requires precise knowledge of the laboratory's skills, know-how and scientific resources. Finally, valorisation benefits from starting upstream of transfer. An inventory of existing scientific capital makes it possible to identify what could be valorised and the collaborations that might follow from it.

This is one of the things the map produced by the ICRI makes possible. At the scale of a university or a SATT, applying it to several laboratories would help identify complementary skills, equipment or know-how useful to other teams. It would also bring to light avenues for scientific collaboration or technology transfer that have not yet been explored. The aim is to gain a better knowledge of each unit's resources, while respecting their own priorities, in order to encourage exchanges and bring new opportunities to the surface.

The limits, in all honesty

Several caveats must nevertheless be kept in mind.

We worked on the data Philippe Drobinski had at the end of his term, data that were neither confidential nor personal. The result therefore describes the laboratory at a given point in time, not its current situation. Some information was not tracked internally; we flagged these gaps rather than giving them a score. The LMD is, moreover, a large unit, well organised and part of a renowned institute. What we observed there cannot be generalised to all public laboratories.

The test also showed what we needed to improve in our own tool. Three months of exchanges were needed to adjust our reading of public research. Feedback from other academic beta testers also helped us to continue this work.

The thermometer and the model

At the LMD, researchers know that a single measurement is not enough to understand a phenomenon. A thermometer shows a temperature; predicting how it will change requires taking many other parameters and their interactions into account.

The same goes for assessing a laboratory's untapped valorisation potential. Publications, contracts and patents provide useful but partial information. With the ICRI, we try to complement them by examining the relationships between skills, resources, partnerships and prospects for development.

After all, a laboratory that studies the mechanisms of climate may also benefit from exploring its own valorisation potential. Some resources, although very much present, sometimes only need to be identified to find new applications.