Explore the National Pilot Interview from France to discover the latest updates on the OSTrails pilot studies. Dive into their national activities and learn about their progress in integrating open science with research assessment. This month, we had the pleasure of speaking with Laurent Romary and Maud Medves from Inria.

- Laurent Romary (Inria, Scientific Information and Culture)	- Maud Medves (Inria, Scientific Information and Culture Direction, Scientific Information Department)

"Integrating standardised software references in machine-actionable DMPs is an important way of signalling how important software is for the creation, transformation or visualisation of research data. Beyond digital science, we aim at generalising such practice to all scholarly fields."

-Can you briefly introduce your organisation? How does it contribute to EOSC? 

Inria is the French national research institute for digital science and technology and gathers over 220 research teams in various scientific fields ranging from digital health to high performance computing. It focuses on research activities in digital science but also leads open science activities (e.g., having pioneered OS software infrastructure by co-founding Software Heritage , taking part in national initiatives on open science in France, development of the French OS monitor). 

Inria is the EOSC mandated organisation for France and is involved in several EOSC projects such as FAIRCORE4EOSC, FAIR-IMPACT and GraspOS. 

As the mandated organisation, Inria takes an active part within the French network of EOSC members, under the auspices of the French Ministry for Higher Education of Research, with the aim of bringing forward, as part of the French contribution to EOSC, the various national infrastructures available in the domains of publication, data and software. 

-What are you most excited about in OSTrails? What are you looking forward to? 

OSTrails is an exciting opportunity for us to improve the way we handle DMP as a component of the research data life cycle. Writing a DMP is often perceived as an administrative burden by project investigators. We do hope that introducing machine-actionable DMPs which can be, to some extent, the basis for further documentations associated with research digital objects (e.g. when deposited in open archives), will be a huge step forward for researchers and their institutions.  

We also look forward to including in maDMPs the description of source code or software that helped creating or handling the corresponding datasets. This is very important in our community, in which software is a central research output. We also hope that it will be an incentive for researchers to better describe their source code and software in open archives, rather than just store them in a git repository. In this respect, Software Heritage, with its systematic archiving of all open-source software code and association of persistent intrinsic identifiers (aka SWHId), provides a solid background for making such references in maDMPs.  

The project will also help us define an institutional policy at Inria concerning research data management at large. 

-How is planning, tracking and assessing research being realised in your country? 

Planning: research planning is mainly addressed at Inria through the support to researchers in writing Data Management Plans (DMP). In France the widely used DMP writing tool is DMP Opidor, which is used by most universities and research performing organisations. Inria is no exception. This online tool allows each institution to customize institutional recommendations for their researchers; support staff can also have access to DMPs drafted by their researchers and provide assistance in enhancing them. 

Tracking: France has the chance of having a unique national publication repository (HAL) available to all institutions for recording their research outputs. Inria has developed a specific policy with a publication mandate in HAL which we are planning to extend to make sure we connect publications with datasets and software. At national level we also track our open science performance in the context of the French Open Science Monitor.  

Assessing: at Inria, the RDM unit staff manually reviews DMPs on request before submission to funders. Funders, whether national (ANR) or European (European Commission), usually perform simple checks of DMPs. At Inria, the Evaluation Committee (the body responsible for assessing the level of excellence of the research conducted at the institute and guaranteeing the level of its recruitment and internal promotions) takes into account work related to software development when conducting annual assessment campaigns. As such Inria policy is fully in based with the direction taken by COARA at international level. 

-What is your pilot about? Can you provide some details on the main actors, services and priorities? How will the results of OSTrails be adopted by your pilot? 

Inria is the unique actor of the French pilot. Though we are focussing on a very specific aspect (source code and software), we will be working closely with partners involved in the French ecosystem of Recherche Data Gouv (our national data repository). Partners of the RDM geographical clusters (Ateliers de la Donnée) as well as the Opidor resource center (Centre de ressources outils communs aux ateliers) will be regularly updated on the project’s progress so that we can integrate OS Trails outputs in the French national tool DMP Opidor.

-Ongoing activities and Next Steps?   

We are currently focussing on ensuring the coherence between activities conducted on machine-actionable DMPs and international standards used to describe software, so that a software component can be introduced smoothly in maDMPs. Such references to software within a maDMP should be expressed in a way compatible with the international software description standard CodeMeta, so that one can directly point to the corresponding entry in Software Heritage.

Thank you both!