Paths to success in Open Science at ICRI 2016.

The International Conference on Research Infrastructures (ICRI) was held in Cape Town in October 2016.

This event brought together several major players in the European and South African research infrastructures, including e-Infrastructures. Apart from the main conference there were several side events which sought to provide a space for discussion and engagement.

The Sci-GaIA consortium convened one of these side events at the Centre for High-Performance Computing (CHPC), with the title “Paths to Success in Open Science in Africa“. The aim of the side event was to bring together members of infrastructure providers to discuss what they are building, and how. Questions such as how Africa could engage with the roadmap of the European Open Science Cloud, or how African e-Infrastructures are being developed to satisfy the needs of research communities, or accommodate the need to share and inter-operate with peer infrastructures abroad were discussed. Rather than tackle high-level policy issues, the workshop aimed to discuss current issues in integration and inter-operation.

The workshop was attended by representatives of funding agencies, EC projects, national and regional e-infrastructure projects and big science projects. Various presentations on the state of the art were made by South African e-Infrastructure providers. The research infrastructure “Commons” was well represented, with presentations of the South African Federated Identity for Research and Education (SAFIRE), and the National Research Foundation. These presented the efforts to develop national identity federations providing better access to services, and the Open Access Mandate, which put forward the case for the development of Open Access repositories and data federations. Platforms in service of research communities such as the African Research Cloud (ARC) and the EGI FedCloud were presented, giving the state of the art in distributed computing infrastructures. On the more technical side, “infrastructure as code” was the theme of a presentation by a representative of the ALICE experiment from the CHPC, describing efforts to make complex services reproducible by adopting a DevOps paradigm.

In the context of Sci-GaIA, these were considered in their relationship to the Africa-Arabia Regional Operations Centre, which acts as a point of coordination and inter-operation between African and European computing infrastructures. The Sci-GaIA Open Science Platform was also discussed, as a model for the integration of all of the tools necessary to enable the execution of Open Science workflows.

From a different point of view, research communities gave their perspective on the _usage_ of these facilities and the challenges faced in exploiting them. Amongst these were the H3ABioNet project of the Human Heredity and Health (H3A) project, and the VI-SEEM project. The perspective of the individual researcher when considering these platforms, and the Open Science paradigm in particular, was provided by Sci-GaIA, highlighting the benefits to be gained from a truly inter-operable e-Infrastructure. Here, the case for Open Infrastructures was made from an “Open Scientist” point of view.

This workshop brought the work that Sci-GaIA is doing in the support of research communities as well as the coordination and strengthening of e-Infrastructure services to the forefront. By bringing together concrete and current examples, we were able to discuss particular points of interest in supporting research communities and better engaging with peer infrastructure. The presentations also helped to inform the upcoming deliverable D3.1 – “E-Infrastructure Sentinel Report”.

E-Infrastructures serve communities best when they are frictionless and transparent. However they are inevitably built by different technical communities with varying priorities, technology stacks, and processes. An Open Science paradigm further highlights the need for the technical communities which build these infrastructures to collaborate effectively. The goal of the workshop was to propose building Open Infrastructures as paths to Open Science. While this goal could hardly be fully achieved in a day’s session, the opportunity for infrastructure developers to discuss thorny topics was invaluable and this work continues.


Dr. Bruce Becker

CSIR – Pretoria, South Africa