Discussions

Consultation on Work Programme 2018-2020 - INFRAEOSC-03-2020

Georgia Tzenou

9 March 2018 - updated 3 years ago
10 comments
Total votes: 11

The European Open Science Cloud (EOSC), presented in the Communication on European Cloud Initiative - Building a competitive data and knowledge economy in Europe, as a part of the Digital Single Market Strategy, aims to give Europe a global lead in scientific data infrastructures and ensure that European scientists reap the full benefits of data-driven science. EOSC will operate as a trusted, open environment for the scientific community for storing, sharing and re-using scientific data and results, supported by high-capacity cloud solutions with super-computing capacity via a European Data Infrastructure.

Horizon 2020 supports the implementation of the EOSC through Work Programmes 2016-2017 & 2018-2020. Some relevant topics include:

European Open Science Cloud for Research (INFRADEV-04-2016);
Developing EOSC, as a secure, permanent, on-demand service-driven, privacy-compliant and sustainable environment integrating/federating e-infrastructure platforms, distributed databases, computing resources and software (EINFRA-12-2017);
Prototyping innovative e-infrastructure platforms and services for universal discoverability of data objects and provenance, and computing with extreme large datasets (EINFRA-21-2017);
Providing access to commercial services through the EOSC hub, based on the aggregated demand of the scientific community, achieving efficiency and economies of scale (INFRAEOSC-01-2018);
Prototyping new innovative services that address relevant aspects of the research data cycle (INFRAEOSC-02-2019).

In order to finalise the last EOSC-related topic of Work Programme 2018-2020 - INFRAEOSC-03-2020, it is important to identify which activities should Horizon 2020 fund during the period 2020-2023 to consolidate and further develop the results achieved through previous topics.

The deadline for this consultation is extended to 6 April 2018.

Please send us your contributions by replying below.

Tags:

Comments

Natalia MANOLA,

06/04/2018 21:01

This response is on behalf of the OpenAIRE consortium.

INFRAEOSC-03-2020 will be the bridge between the current and pending H2020 programmes and the co-design, co-creation and mission-oriented research proposed in FP9. Realizing that there is a stack of service layers addressing different needs of the data life cycle which need different approaches by different smaller and focused projects, and that the current programme is only just starting, we provide a list of concrete actions and priorities for the next phase:

Supporting services with a user-centered approach. The voice of researchers is the one indispensable prerequisite of user-centered design and uptake. Representation of researchers in the design and use of e-Infrastructures is a necessary but not a sufficient condition to implement user-centered services. In the everyday operations and in the innovation planning of EOSC, researchers will require local support to learn about the use of e-Infrastructures/EOSC as well as specify and express their demands.

A day-to-day mechanism for researchers to utter demands regarding the EOSC should be developed on the basis of local support. Mixing RIs or thematic domains and e-Infrastructures is a first step forward, but it is essential to involve the long tail of science stakeholders (RPOs, libraries, SMEs).
Give emphasis to the uptake of EOSC services by ERC and Marie Curie researchers.
Current programme focuses on service provision KPIs. The next programme must additionally focus on user-centric KPIs using industry based measurement tools and standards.

Strengthen the Open Science as-a-Service approach. Research infrastructures are “vertical endeavors” that should focus on devising thematic services for science, while relying on e-infrastructure “horizontal endeavors” to benefit from network, storage, computation, and publishing functionalities. The degree of support and integration of horizontal into vertical must be seen as an opportunity not as an obligation. Horizontal services should be packaged in order to be transparently and seamlessly provided as-a-Service to research infrastructures according to their specific needs.

Put software in the research agenda. To truly enable open science, we must recognize software as a key research outcome that facilitates transparency and reproducibility. Put the infrastructure in place that will support guidelines and good practices, measures to support communities to change to these, links to EOSC, RI and national services. Any foreseen efforts must take account international initiatives and join efforts with relevant scholarly societies (e.g., ACM, IEEE) and the industry.

Develop and operate an Open Metrics infrastructure. To reach the bulk of researchers as the main users of EOSC, a language of scholarly communication should be spoken. Excellent researchers do not have time to design infrastructure or temper technology but are eventually targeted towards communicating with peers through publishing literature and other forms of research results, which contribute to their career development. EOSC, enabling the Next Generation Metrics, will require an infrastructure of open metrics which will define and maintain formats, exchange protocols, guidelines on the use of metrics and indicators, and will host data related to metrics making them openly available to all interested parties.

Make training services for data skills as well as open science skills an integral part of any infrastructure, ensuring the inclusion of ethical and privacy issues (GDPR). Ensure that INFRAEOSC-03 will develop aligned efforts amongst key infrastructures and RIs, realizing the different levels of EOSC layers and stakeholder approaches. Anticipate the use of resources providing ad hoc support where necessary for the use of EOSC.

Strengthen the support for open science policy development and alignment at national levels. Develop and establish the tools and services so that national funders may easily align their policies, monitoring and decision making processes. These tools may include: common formats and structures for sharing their policies and their public funding information (towards an EU CRIS), as well as the use/best practices of standardized DMP and GDPR mechanisms and protocols (e.g., checklists).

Support the adoption of alternative business models for soft services. Procurement has been instrumental in changing the network and computing service provision landscape. Extend (pilot) this or similar mechanism beyond network/hardware services to cover services related to scholarly communication, publishing or analytics, e.g., PIDs, subscriptions/APCs, common data analysis/visualization services.

Make scholarly communication more intelligent. Facilitate the use of text and data mining and AI to build innovative services to foster the re-use of open research results (publications and data) via automated extraction, combination, and analysis. Horizontal and collaborative efforts like OpenMinTeD (sharing, re-using content and services, as well as a sandbox for innovative services) are key towards achieving the co-design and user-centric efforts.

Widen the scope of services targeting further than research communities. Facilitate the mechanisms for crossover between research infrastructures and other application domains beyond academia, to create research infrastructures more capable of answering the large-scale and multifaceted societal problems around immigration, wellbeing and urban and economic regeneration in areas of industrial transition. Examples are development of TDM techniques from the humanities for use on legal texts, linking research to public health policy reports by governments and supranational bodies, integrate transport research in urban planning. Supporting Open innovation and digital innovation hubs are key into achieving this goal.

Establish continuous and transparent monitoring mechanisms. Reinforce existing efforts of service and data catalogues and expand with other types of product catalogues. Support infrastructures to align their KPIs on production, operation (include costs) and uptake (include user centric KPIs) of their products.

Target international collaborations for greater interoperability of scholarly communication/ data infrastructures, including open and FAIR of all research results. Of particular importance is the support of Europe as the front runner at implementing on a global scale COAR’s Next Generation Repositories, some aspects already started in OpenAIRE-Advance.

Hendrik Ike,

06/04/2018 19:11

GÉANT response to Consultation on Work Programme 2018-2020 - INFRAEOSC-03-2020

The integration and consolidation of pan-European access mechanisms to public e-Infrastructures & commercial services through the EOSC hub provides the opportunity to bridge the gap between the services offered by Europe’s e-Infrastructures within the EOSC hub and the research community.

This call will contribute to the enlargement, deepening and widening of EOSC. All ERA and EHEA associated countries should also have the opportunity to be involved. The nature of the research data to be accessed through the project should also be widened, for example utilising Earth Observation data project (OCRE) as a prototype service that will be further evaluated and developed. This includes the OCRE cloud services procurement which would also have to be integrated. This will truly capture the proven EU-added value of such a project. The integration of access must therefore also be designed with scalability, ease of use and fitness for purpose in mind. There must be agreement on a federated architecture for authentication and authorisation, allowing for the incorporation of existing products and services. This will reduce duplication of effort, will leverage existing widely developed authentication services like eduGAIN and will allow research collaborations to select those authorisation mechanisms that best fit their needs without sacrificing interoperability. That way a healthy ecosystem is created with fair competition between access services and in which novel approaches and new tools can be incorporated, as long as they adhere to the common AAI architecture.

In order for INFRAEOSC-03 to be a success, access mechanisms should not be designed on a one ‘single point of access’ model for users. Separate e-Infrastructures have developed their own access mechanisms and an intermediate period of testing a federated access model is advantageous both for users and providers of data. In case the single point of access will be supported, it is critical that the model chosen is inclusive, support researchers including universities over ERA and EHEA. The project has to build the trust, engage all partners and develop the necessary skills-sets. Special attention also has to be paid to gain public support.

It is essential to integrate services for scalable computing, software, distributed databases and storage with services for access to and preservation of scientific information across Europe. The GÉANT network and services portfolio provides a strong platform which already connects over 50 million users at 10,000 institutions across Europe and supporting all scientific disciplines. In order for such a multi-faceted system of research and provision to function, federated mechanisms of access will deserve a similar method of sustained funding support and the development of skills and training in order for pan-European access mechanisms to be efficient. The ongoing transition to develop European pre-exascale and exascale computing and maximize efficiency and effectiveness of existing and future e-infrastructures at MS and EU level has to remain a key driver to secure scalability and sustainability.

Finally, the project will need to incorporate the latest legislation from the EC work plan, with full attention and compliance paid to GDPR and Free Flow of Data, Public Sector Information, etc.

Per Öster,

06/04/2018 15:56

INFRAEOSC-03-2020 must support multidisciplinary research, addressing the issue how to combine and reuse data from different sources. It must support the EOSC governance to put strong incentives in place to promote collaboration among existing institutions and frameworks.

Data management and storing in EOSC will follow the FAIR principles. However, these do not address for example data structure, metadata standards, transformability of the data, and its reuse. Data integration does not always equal to data interoperability. For example, genetic data is not reusable without specific tools. Thus, there has to be semantic interoperability (e.g. controlled ontologies and vocabularies, file formats) in addition to standardised technical protocols and interfaces. Obviously, research communities need to be deeply involved in implementation of EOSC to eventually be able combine and reuse data, e.g. do data integration and data fusion for new purposes with new tools.

EOSC should promote ethical use of data in all its activities and throughout the data lifecycle, especially when data is integrated and fused with other data sets. Licensing in data fusion requires deliberate principles and processes. Data tags could be developed e.g. for automatic recognition of sensitive data. Regulation by the GDPR must be taken into account when combining and transferring data. With possibility to automatically compile metadata, data would also be easier to make findable and reusable. This would be a major opportunity for AI development, innovation and new services.

New incentives for researchers to produce, describe, open, combine, and reuse high-quality data are required. Rewards for describing datasets for opening and reusing them by humans and machines (AI) must be created. Promoting commercial use of scientific data is important for society as it brings value and makes public investment more profitable. Data and services should be available also for commercial users. New services should be developed e.g to facilitate data location and transfer of datasets as objects, not files. Standards for long-term preservation of data should be defined to enable, for example, data transfer from a repository to another and enable FAIR preserved data. Finally, issues such as, how end-users can discover services, can services be moved, and can status of services be identified, should be addressed.

Specifically CSC suggests that INFRAEOSC-03-2020 should

ensure agility by allocate resources to several smaller projects instead of few large projects to be able to adjust to new needs;
support the full EOSC eco-systems with focus on increasing the number of compliant and compatible data resources and services while keeping the governance, technical and operational coordination lean and fit-for purpose;
focus on projects that promote joint efforts of research communities, research infrastructures, and e-infrastructures to fully benefit from tools and services provided by EOSC;

attempt to develop policies and business models around scientific data
create incentives for multidisciplinary research combining, knowledge, tools, resources, and efforts from both research communities and research infrastructures;
support service providers to develop their interfaces, and in their service development projects in terms of e.g. multidisciplinary data interoperability.

For CSC - IT Center for Science Ltd

Odile Hologne,

30/03/2018 16:11

Agri-food science are complex from a data science point of view, with different disciplines (from genomics to social sciences), different scales (from genes to ecosystems) and geolocalisation. The ability to integrate these heterogeneous data is a key issue to tackle new societal challenges (e.g. food security in the face of climate change, sustainable food value chains, digital agriculture and food technology). In addition, the automation of data collection, new techniques in “omics” as well as the development of new types of data sources (e.g. Internet of Things, crowd-sourcing, text mining) has allowed to collect an exponentially increasing amount of data. Thus, there is a need for an e-infrastructure that connects data, infrastructures, resources and people and that allows to share efforts and expertise, and support innovation. In particular, such an e-infrastructure can develop Virtual Research Environments (VREs) that provide tailored solutions for specific communities.

Based on the work done in the e-ROSA project http://erosa.aginfra.eu, we consider that the next call should focus on :

- the interoperability of Digital objects using semantic resources and technologies, creating semantic infrastructures that support the interoperability and reusability of data

- Creating competence centers for capacity development to accompany the implementation

- involving researchers in information technology to develop new services based on cutting edge technologies

Alasdair Reid,

30/03/2018 15:01

A response to the consultation by EFIS Centre, a Coordinator of the Horizon 2020 funded eInfraCentral project.

Many of the projects which activities and more importantly results can be of use for the finalisation of the last EOSC-related topic are still running. We welcome this consultation and would like to share some thoughts from our ongoing project eInfraCentral: http://einfracentral.eu

Disclosure: This response presents the views of EFIS, as a Project Coordinator, and is not a response on behalf of the whole consortium.

Standardisation

To ensure a smooth and user-centric approach in accessing various services provided by different type of infrastructures, from publicly funded e-infrastructure service, ESFRI RIs, national RIs to private service providers, it is crucial to use a common grounds in how the EOSC services will be accessed are being presented. Development of an ISO-type standard for EOSC catalogues of services can be one step in ensuring the efficiency of platform and service provision. The work done under the eInfraCentral project in developing a common catalogue of services and Standard Template Description provide a basis for the development of such a standard. The work was done taking on board previous efforts. Standardisation can also help with automated data mining from the websites of individual research infrastructures and service providers, thus adding to the efficiency.

Aligning national and EU-wide efforts

Integration of the national and EU-level efforts around EOSC as well as services by various types of research infrastructures (not only e-infrastructues) is crucial. It is important to create synergies between EOSC-related efforts at national level (usually nationally funded, although in many cases also EU co-financed) and EU-wide efforts like EOSC-hub.

To this end, a review of legislation/legal frameworks and how national procurement mechanisms help/hamper access of users to services need to take place. A detailed mapping of national e-services or their integration in the eInfraCentral catalogue and the future EOSC Portal will also help to optimise and align user access and use of e-services. Given different funding sources and infrastructures, a joint approach between several Directorates-General of the EC may be needed.

Involvement of private sector

Inclusion of the private service providers is key if we want to keep the user perspective in mind when developing EOSC. Questions of procurement and standardisation will need to be covered here.

Data protection

Provision of access for users might inevitably lead to some of the users’ personal data being collected. The General Data Protection Regulation (GDPR) needs to be taken into account in any further development of EOSC.

Andreas Walli,

29/03/2018 14:03

We welcome your initiative to identify activities needed and potentially funded for the European Open Science Cloud. In our view, the prototyping of a new innovative service that can address current information gaps in the studies of dynamic land processes is particularly suited for INFRAEOSC-02-2019.

Land-use science provides critical advances for better understanding land use in light of global economic and environmental changes. The eminent importance of land-use science for global sustainability therefore attracts substantial attention of academia. For example, land cover and its changes modify the goods and services provided to human society (e.g., the provision of food and fibre, recreational opportunities, etc.), force climate by altering water and energy exchanges with the atmosphere, and change greenhouse gas and aerosol sources and sinks. Land-cover distribution also is partly determined by regional climate, so changes in land cover may indicate climate change. However, most land cover change has an anthropogenic cause with an estimated 42-68% of the global land surface impacted by land use activities (crop, pasture, wood harvest).

Yet, such dynamics processes are to date captured only through static land cover assessment products. The European Open Science Cloud work programme would be highly suitable to enable the capacities needed for serving a science specific, dynamic land-cover assessment service to fulfil the mission of scholarships stemming from geography, economics, sociology, forestry, environmental sciences, climate change and other disciplines that have ‘land’ as a central theme in the study of coupled human–natural systems. Thus we would very welcome if the INFRAEOSC-02-2019 could nominate such an innovative service related to the assessment of dynamic land cover processes.

Bob JONES,

29/03/2018 08:40

The timing of this consultation means it cannot benefit from any insights gathered from the results of the projects funded under the calls cited in the consultation announcement, namely EINFRA-12-2017, EINFRA-21-2017, INFRAEOSC-01-2018 and INFRAEOSC-02-2019.

Even more importantly, this consultation does not offer the opportunity to gather input from the ESFRI cluster projects that will be funded under call INFRAEOSC-04-2018 ‘connecting ESFRI infrastructures through cluster projects’ with the risk that the consultation will focus on a public e-infrastructure service provider point of view rather than how it serves these major flagship and landmark research undertakings.

Taking the above situation into account, we would like to make the following remarks:

The use of commercial cloud services continues to expand, so we see the need for further support for the integration of commercial services into the EOSC and means to simplify their uptake. This includes simplified procurement processes for procurers and the end-users they sponsor, as well as the harmonisation of SLAs and contractual terms and conditions for the engagement of conformant service providers into the EOSC.
We are experiencing a diversification in the range computing architectures being used by research communities and this need to be accommodated in the mechanisms by which external services can be integrated with EOSC. This includes expanded support for the integration of ESFRI Research Infrastructures and interoperability between EOSC and EDI.
As the implementation of EOSC proceeds and support for F.A.I.R and open access grows, it will be necessary to make available data preservation services that can ensure the continued access to and re-use of publicly funded datasets in the long-term (decades).
Recognise the importance of ERC grantees and Marie-Curie research projects as potential user groups that could make use of EOSC services.
Ensure EOSC can accelerate innovation through the re-use of publicly funded research data by offering incentives for the development of data-intensive services (by the private and public sector) in the context of the platform
Foresee training services for researchers and service providers wishing to make use of EOSC.

Tiziana Ferrari,

29/03/2018 05:33

EGI response to consultation: the EOSC Data Commons

INFRAEOSC-03 should fund an innovative data and compute integrated infrastructure - the EOSC Data Commons - that through the federation of research data, computing, applications and other open science resources, responds to the problem of scalable access to research data through a dedicated data provisioning service approach that is complementary to the traditional data download model.

The EOSC Data Commons should:

Federate services, data and resources (from e-Infrastructures, Research Infrastructures and other providers of European relevance), and ensure their coordinated procurement, management and provisioning.
Address the need of accessing research data and data products close to processing facilities in a scalable way while avoiding duplication of data storage activities across Europe, in order to allow further value extraction from the data, increase research data exploitation, and reduce costs.
Provide research data as a service with the active participation of Research Infrastructures, and deliver the data close to processing facilities. Through this, the infrastructure enables the creation and provisioning of added-value services. By doing so, the EOSC Data Commons will be an innovative European platforms for all RIs on which their research data of public relevance can be exploited.
Allow to discover, search and access major public research data sets and information, and to access different processing capabilities. A co-located distributed cloud infrastructure will offer the possibility to run a library of scientific applications, to share these across applications, and to offer them as Software as a Service.
Allow the federation and integration of publicly funded computing resources across Europe to support the needs of European researchers both at the level of computing capacity and capabilities (e.g. HTC, Cloud) to ensure that researchers have access to geographically distributed large capacity computing resources and specialized computing facilities including accelerated computing and experimental technologies. The infrastructure should maximize the return of European and national investments in publicly funded computing e-infrastructures by enabling shared open access via federation of distributed computing resources.

The data component - funded by the EC to ensure research data is broadly accessible as a public good - should be complemented by compute, application and other open science resources funded by national funding agencies and research performing organizations.

The implementation of the EOSC Data Commons will require:

Continual support to the federated service management processes, activities, policies and supporting tools that realize the federation of national, regional and digital research infrastructures of European and international relevance.
Access provisioning services to enable the secure access to research data, resources and services, including an integrated and federated fabric of identity provisioning, authentication and authorization services, in order to lower barriers in federated service access.
Integrating activities involving e-Infrastructures and Research Infrastructures, to: (1) integrate services; (2) foster the adoption of common interfaces, standards and best practices for service interoperability; (3) develop and expand a joint service strategy and maintain a common service catalogue; (4) co-develop new services and customize existing ones.
The development of skills and human networks through a network of Competency Centres involving expert users, technology providers, service providers, industry and SMEs.
International e-Infrastructure cooperation to support international research projects and communities.
The funding of access provisioning and marginal costs incurred when enabling international access to open science resources and services of global relevance.

The EOSC Data Commons are intended to be a joint endeavor of Research Communities, Research Infrastructure projects and landmarks, e-Infrastructures and other providers of European and international relevance, and will contribute to the realization of the EOSC vision.

Georgios Megas,

19/03/2018 14:50

Towards EC's goal of including the private sector as well, additional funding could be offered ad hoc to close to markets solutions from SME Instrument or FTI that could utilize or offer data, meta data, results etc from the Open Cloud.

Include specific calls and topics related to cross-sectorial and cross-lingual data integration and experimentation (the title is borrowed from a highly important H2020 ICT Topic of 2016: http://ec.europa.eu/research/participants/portal/desktop/en/opportunities/h2020/topics/ict-14-2016-2017.html

New Business model Creation maybe through "Geant" etc.

A major issue would be security to assure the safety of the data and the owners even if those data are open.

Funding for dissemination of the offered services will be needed.

Specific funding should be foreseen for the creation of a network that would provide information regarding the IP protection and the closing or opening of data and the legal procedures for inclusing the private industry in the initiative. Similar to INFRAEOSC-01-2018.

Modification of the interface and continuous improvement would need funding.

Support for Cloud computing migration might be needed for some organisations and even companies.

Rule creation and clarification regarding private procurement to the EC regarding cloud infrastractures and O&M. Re evaluate the already established (in 2018 or 2019) procurement mechanism.

New Hyper Computing technologies would have arise that would be in lower or higher TRLs in the near future, thus calls or topics that would try to implement such technologies and make the cloud more reliable, more efficent and fast and more secure should be foreseen.

Calls that could include non curated data right out of IoT activities that could help reasearch somehow should be taken into account whenever its possible.

Inclusion of FSTP mechanisms to engage SMEs in the provision of solutions should be foreseen as well.

G.M.

Digital4Science

ARCHIVED PLATFORM

This website is archived since mid-May 2021 and it is not possible to login anymore.

But don't worry there is a new Futurium platform and all active communities have been migrated.

Discussions

Consultation on Work Programme 2018-2020 - INFRAEOSC-03-2020

Comments

ARCHIVED PLATFORM

This website is archived since mid-May 2021 and it is not possible to login anymore.

But don't worry there is a new Futurium platform and all active communities have been migrated.

Share this:

Comments