ESFRI Workshop: “EOSC and ESFRI RIs services: how do they tango?”
London, 30 January 2019
Following the Competitiveness Council Conclusions of May 2019, which recommends to establish an effective coordination between ESFRI and EOSC, ESRFI has invited EOSC stakeholders and ESFRI cluster projects under the motto “EOSC and ESFRI RIs services: how do they tango?”. In the first session of the workshop the EOSC-related projects EOSC-hub, eInfraCentral, FREYA, OpenAIRE-Advance and EOSC Secretariat presented horizontal services provided to the vertical, domain-specific research communities. The research communities themselves were represented in the second session by the ENVRI-FAIR, EOSC-LIFE, ESCAPE, PANOSC and SSHOC. Parallel sessions addressed four main areas of interest;
- Good practices and business models,
- Open Science, Fair data, Reproducibility (Reliability),
- Architecture: technologies, platforms, interfaces, and
- Skills and training.
Good practices and business models
In the session Good Practices and Business Models it was said that the EOSC should be adapted to the existing conditions and existing investment by Member States and Associated Countries’ in data infrastructures. The EOSC must take the existing boundaries into account. Experiences in different RIs show that there is already a good use of the existing generic e-Infrastructures. With regard to the generation of data a given heterogenity is recognised, e.g. data from seismological, vulcanological geo-resource source could not easily transferred to EOSC. Furthermore, the aspect of ownership of some data sets needs to be addressed before these data could be accessed within and across scientific domains. Another topic discussed in this session was around performance indicators and the need to create proper KPIs. It was proposed that the development of KPIs not only address that data is shared from the community to the EOSC, but also that these data are used by a broader community. Furthermore, it was commented that business model(s) of the EOSC must be connected to the concept of the Commons and that conveying the EOSC to national investments is important.
Open Science, FAIR data, Reproducibility (Reliability)
In the session Open Science, FAIR data, Reproducibility (Reliability) the consequences and challenges of implementing the FAIR principles were discussed. The speakers highlighted the following challenges:
- an issue of engagement with the research community,
- a need for open feedback loops on data re-use and re-purposing.
The costs of not having FAIR data of at least 10bn EUR have been estimated and also the considerable investments were mentioned. The case was being made by RIs for the investments needed to implementing the FAIR principles. It is expected that EOSC in particular will help in terms of efficiency and opportunities to align FAIR and Open Science requirements.
Other challenges like the alignment between clusters and across clusters, the setting of embargos for data re-use within or outside a RI, issues around data quality, need for feedback loops from the user, the research community to the data providers, issue on how to deal with sensitive data and to what extend that is being dealt within EOSC have been indicated in the session.
Furthermore, the need to demonstrate measurable impact was addressed in the session and the need to identify and develop transparent metrics, which help RIs to show their value and allow them to quantify to their funders and their communities the impact they have.
Architecture: technologies, platforms, interfaces
In the session Architecture: technologies, platforms, interfaces the discussion outlined that AAI is a much-needed service in data infrastructures. The interoperability of data provides open questions due to unclear borders between domains and it is important how to interface. The EOSC is expected to provide guidelines and populate open data sets. It was also said in the session that many projects have FAIR data in their value proposition but that it is hard to and unclear how to get a grip on those that are useful. More so, if the EOSC is not going to provide solutions to these questions, Google will do. It is thus of utmost importance for the scientific community to provide solutions that address the users’ needs.
Skills and training
The main discussion in the Skills and training session was around the topic of sustainability, implementation and inclusiveness. It was highlighted that the EOSC is an endeavour of inclusiveness. Many EOSC projects have a different timing, and thus an increased need for coordination is recognised to maximise the outcome. Furthermore, the exchange of best practice and the professionalisation the training efforts was highlighted.
In the final session a discussion about the EOSC portal sparked the interest of the audience. Many attendees pointed out that there will be more than one entry point to the EOSC, which then will be more dynamic and connecting the demands with the offerings. It is expected that these entry points will enable data sharing in an automated fashion and will be not only browsable platforms but rather a way to connect the human networks and the experts. A first step towards this dynamic approach could be the provision of a set of APIs for everything that works.
Jan Hrušák, Chair of EFSRI, concluded that the EOSC is not a revolution but rather an evolutionary process, which will help to facilitate this cultural change.
The presentations and videos of the sessions are available here: https://www.esfri.eu/esfri-events/esfri-ris-eosc-liaison-workshop