Increasingly, research must handle more complex data at higher rates of production and consumption. Researchers require increasing computing power and sophisticated tools to help share, process and analyse data, across all disciplines – not just those traditionally associated with high-performance supercomputing.
Improved digital infrastructures are key to supporting researchers in their work. Cloud-based resources are an important element of these infrastructures and provide benefits such as increased collaboration. For instance, if multiple researchers from different institutions or even different countries are working on the same project, access to shared resources is essential. When resources are stored on a cloud platform, researchers can access these resources through their institution’s credentials, from wherever they are in the world.
European Open Science Cloud
The European Open Science Cloud (EOSC) has the vision of providing a cloud platform to offer 1.7 million European researchers and 70 million professionals across science, technology, the humanities and social sciences an open service for storage, data management, analysis and reuse of research data.
The EOSC was first proposed in 2016 as part of the European Cloud Initiative to build a ‘competitive data and knowledge economy in Europe. The aim of the EOSC is to work across borders and disciplines by federating existing data infrastructures dispersed across the EU, and will be a key pillar of the new European Data Strategy. Jisc, alongside many other partners across Europe, has been participating in the projects and working groups defining the architecture and governance of EOSC.
This approach to open research is proving particularly useful during the current climate: lead by the European Bioinformatics Institute (EMBL-EBI), the European Commission has launched European COVID-19 Data Platform explicitly for Coronavirus (COVID-19) research, which will be a key component of the EOSC.
The platform contains data on coronaviruses, including samples from COVID-19, which is regularly updated from international databanks. The cloud platform also provides some tools to aid identification and analysis of the samples, which means that researchers can run experiments virtually, without having to install these tools on their own machines.
Proceed with caution
As various providers are offering more tools as Software-as-a-Service (SaaS), the role of cloud is expanding beyond that of just computing and storage solutions. Researchers can explore and experiment with new tools, technologies and computing platforms quickly without large upfront financial and infrastructure costs. That said, as with all digital tools, the benefits are dependent on how they are used.
A certain amount of caution is needed in the case of pre-built toolkits, as they can also be misused or introduce unexpected biases. If a researcher knows how to use the toolkits and understands the parameters and limitations of the technology in question – for example, AI – they can be really useful. However, if this isn’t the case and the tools aren’t used properly, results could be compromised, for example, mistaking correlation for causation, or misinterpreting artefacts inherent in the tools as trends in the underlying data. The “black-box” nature of some of these SaaS tools - whereby their internal workings remain hidden - introduces challenges for effective peer review. While one of the advantages promoted for SaaS is that software updates are managed for users transparently it can impact the reproducibility of research results.
However, when used well, SaaS tools can improve the flexibility and efficiency of research projects. For example, if a researcher has an idea for an experiment but is unsure whether it will work, they can spin up a machine in the cloud instead of having to order high-performance computing (HPC) kit, which can be both expensive and time-consuming. When the experiment has been modelled in the cloud, however, it can be more cost-effective to use dedicated HPC for ongoing projects. Working with accredited partners such as Jisc can also help make the most effective use of cloud services including AWS and Microsoft, further extending budgets.
It’s all about balance
Increased availability of cloud computing for research provides options for different projects, and in the current health crisis, it is proving an important tool in global vaccine efforts. This is evident here in the UK through projects such as the University of Bristol’s use of cloud to develop the synthetic ADDomer vaccine.
The global challenge brought about by Coronavirus has shed greater light on cloud’s place in research collaboration, and as data analysis continues to take centre stage in research, we will only see cloud usage grow.
For more information on cloud solutions for education and research, visit jisc.ac.uk/cloud
