The EOSC Symposium 2022 marked the return to in-person events for the EOSC community. And similar to a child or an adolescent, the growth and increased maturity is much more apparent when not limited to a Zoom window. During the last three years, EOSC has effectively moved out of its “parents’ house” and established a new, tripartite family.

For those of us who are navigating the challenges of implementing EOSC it is worth prefacing this overview with the observations of our Canadian RDA colleague, Mark Leggott, from the Digital Research Alliance of Canada, who observed: 

“It is a prescient decision of the EC to fund EOSC, from the outside it has developed hugely in the last few years, and from other global perspectives it is pushing forward critical infrastructure features for open science infrastructure and services, and from my perspective, the highlight to it all is the development of the AAI framework.” Mark Leggott, Director of International Relations, Digital Research Alliance of Canada

Coming of Age 

You could see signs of this coming of age everywhere, from governance to technical architecture. EC Keynote message at the Symposium was noted in the Commissioners’ calendar (similar to COP27 participation). ESFRI Research Infrastructures are now an integral part of the EOSC community, and the borders between individual e-Infrastructures have practically disappeared from technical and organisational presentations. It seems that an “Open Science Schengen Convention” has been ratified and implemented by the community. 

Not that EOSC admits it is “ready”. The biggest EOSC-related event in the calendar, the four-day event was a whirlwind of projects, workshops, heated discussions proving that it has a long way to realise the scope and level of ambition different components of the ecosystem should aim at. Long-term aspects and FAIR - including software and services supporting FAIR and reproducible research - were discussed in, and out, of the sessions. 

In general, change is easy only when viewed sufficiently far in the future. The road ahead is not always clear, and it is known that some parts need to be built over uncharted terrain. However, if you needed to condense the symposium into one word, the best one would probably be: “Smooth”.

RDA and EOSC 

RDA brings the disciplinary community connection to EOSC, and its role in the EOSC Future project, RDA was present in a number of sessions including the Researcher Engagement session. Beth Knazook of the DRI presented the work of RDA engaging disciplinary communities to EOSC, working to understand their needs and readiness for EOSC engagement, via RDA channels, and working with the network of RDA/EOSC Future Domain Ambassadors. Domains such as Material Sciences and Engineering are lagging behind in open science practices whereas life sciences are well-served with infrastructures and data sharing processes. Networking, informal discussions are important mechanisms among underrepresented research domains as are creating a sense of community involvement.

A new RDA WG on also presented their work on AI and Data Visitation in the context of EOSC and the the RDA/EOSC Future  awardee network was very visible. In general, it was good to see the RDA community growing and so active! This is bolstered by the EC continuing support for RDA as mentioned in its Policy Actions for 2023, as mentioned by Michael Arentoft of the EC's Open Science Unit DG R&I.

Some other observations/ main takeaways: 

The following observations and quotes a illustrate the multifaceted nature of the event. EOSC brings together experts with backgrounds in large-scale distributed computing systems, traditional “big research data” (high volume, high velocity, low variability) and FAIR research data relying on complex and metadata and ontology solutions.

How to make EOSC Engaging

• A “Pidgin EOSC” is needed. One of the main take-away’s is the need for clear definitions for key terms. 

• We would need to clarify the foundational concepts and their relationships: EOSC Core, Exchange, Portal and Marketplace. For example, the Service Management System session clarified the practical differences between Core and Exchange services in a relatively concise but comprehensive manner. However, presenting this coherently in the portal/marketplace context needs further attention. 

• Research communities have typically their specific configurations of workflow and analysis tools, as well as practices for using them. This poses a discoverability problem: at the moment, the standard research workflow is unlikely to include search of resources in the EOSC portal. The requirements and assumptions related to specific resource types, such as storage or computing, can also differ quite dramatically between different research domains (as noted in the session focused on the EOSC Execution Framework)

• Engaging via use cases suggested to the EOSC TF on Researcher engagement, a nice example from the Semantic interoperability group. A librarian in the audience made a plea for as many user stories as possible AND that you can show the good and the bad - mistakes are also useful. 

• EOSC can be a playground - in terms of messaging we should emphasise that EOSC can be “playground for data stewards” “ EOSC is about people who are excited about open science and who want to connect, various levels of enthusiasm”. 

Opening up EOSC to the ‘Other’

• We should all be seen as users of EOSC, including Citizen scientists 

  - citizens are core users but also providers of EOSC and the more they engage the richer it becomes. 

• Considering that entry to EOSC is gated, EOSC can have a potentially strong bargaining power towards commercial providers (i.e., on data egress, volume discounts etc.) as it can provide access to thousands of research institutions and more than a million researchers for favourable conditions. 

• ENVRI have also recently implemented an EOSC-powered dashboard, collecting their results all in one place, demonstrating good use of EOSC implementation. 

The user journey to EOSC - ditch the jargon

It was publicly acknowledged the portal needs a 360 degree revamp to make it understandable and usable for the user.   The new platform should be jargon free - for the front end researcher - get rid of terms such as Aggregators and Integrators! After much user testing, The new release of the EOSC portal means you can search publications, data, software, services all from one place. Take a look. 

Below is an overview of what users should be able to do in the portal: 

Squaring the Circle: Horizontal meets Vertical

Exactly how can the disciplinary research infrastructures use the services of the e-Infrastructures and EOSC Core? What services are useful for them? This has been a long running question and the session with ENVRI-FAIR and the e-Infrastructures went a good way to answer this.  ENVRI indicated features they would like to use within their research workflows: Jupyter notebooks, cross domain AAI and Federated searching. 

ENVRI have also recently implemented an EOSC-powered dashboard, collecting their results all in one place, demonstrating good use of EOSC implementation. 

Note! Active help is needed for disciplinary metadata mapping to EOSC - this is a big question and will be the same for all vertical RIs and EOSC. Perhaps a role for RDA here.

Sustainability

The role of a credible, clearly communicated long-term plan as a prerequisite for the growth and trust of the user communities was one of the subtexts of the event. “Sustain or perish” is perhaps a truism, but turning it into a concrete, efficient and equitable sustainability and governance models is a formidable task. Take a look at the EOSC Sustainability TF session. 

Putting ethics into EOSC

The new RDA WG on AI and Data Visitation also led to a discussion which led to the role of AI but also ethics in EOSC. There might well be a need for an Ethics Board or a presence to anticipate AI and Ethics in EOSC. This RDA WG can be useful to guide here. A topic that came up a number of times was the tricky issue of GDPR. Many EC Horizon 2020 projects were set up back in 2016 and while many had ethics boards in place, new GDPR regulations were put in place which was challenging to say the least. If EOSC could tackle this issue much would be gained in terms of uptake and use of EOSC. 

All Open Science counts as a contribution EOSC:  The National Commitment 

Member states are all working towards open science at varying levels. Any success here means a success for EOSC. As a policy tool, EOSC also provides top down pressure to implement open science. For some countries, e.g Norway, national alignment with EOSC has actually initiated a clear path for disciplinary groups to come together and to use generic services.

From technical arrangements to discussions about applicable first principles, the mood seemed to be that everyone needs to and is welcome to join in to define and build the EOSC. And RDA is happy to contribute in bringing its communities on board.  

RDA will continue doing its fair (or FAIR) share in building EOSC. In addition to day-to-day contributions through various projects, RDA will co-organise a dedicated satellite event before the RDA plenary in March. RDA will also naturally join the EOSC symposium that will be organised in September 2023 in Spain (details TBD).

By Matti Heikkurinen  and Najla Rettberg, RDA Europe. November 2022

Photos courtesy of Matti Heikkurinen