Materials Cloud

Swissuniversities P-5 "Materials Cloud" project

Part of the development of Materials Cloud is supported by a Swissuniversities P-5 grant, titled "Materials Cloud as an open, international, and fully FAIR repository of computational data and workflows" (project ID: 182-008), with funding for the 2019-2020 period. The project has also been renewed and receives additional funding in 2021.

The main implementation actions that are addressed within this grant are:

Providing research as open data
Opening up an existing service to other participants, including international participants (including various disciplines like materials science, computational physics, computational chemistry)
Combining existing services into a joint solution that is open to participants (by aiming to integrate with national and European existing services and networks)
Testing and improving the quality of the services and further developing them (by implementing solid, robust and scalable technical solutions: monitoring, backups, autoscaling, ...)
Expanding established, local solutions in research data management, data analysis and data evaluation (by providing DMP templates for researchers, also in collaboration with other projects funded by swissuniversities)
Providing training modules for research data management and eScience (in the form of courses, tutorials, webinars)

The abstract of this grant, describing what Materials Cloud and AiiDA are, and outlining the aim of the project, follows:

Materials Cloud is a web platform for Computational Open Science, conceived to assist Computational Scientists in the full life-cycle of their projects. In general, we can distinguish three types of data: social, harvested and noisy; experimental, harvested in controlled conditions; and computational, generated in controlled conditions. Materials Cloud focuses on the last type. As such it focuses on two essential components:

assist the researchers in the data generation part,
track, visualize, and publish the controlled conditions under which the data was generated, i.e., record and share of the data provenance (the logical chain or, more precisely, graph of all input data and calculations that were used to produce the final result).

Based on these principles, the core of Materials Cloud is a data generation platform, strongly coupled with an archive and a sharing platform for open research data, assisting scientists in data dissemination, curation and publishing with DOIs. This sharing platform, thanks to a rich web interface, allows users to browse the data, navigate its provenance, visualize it, and download it for further reuse, making data fully accessible and interoperable. The platform is powered by our software framework AiiDA, that provides automation of simulations, of workflows and a completely automated provenance tracking to ensure reusability and reproducibility of all computational research. Moreover, AiiDA has already a collection of plugins supporting over 20 simulation packages, and implementing various relevant scientific workflows. Thanks to the unique combination of features of AiiDA and Materials Cloud, our platform can ensure FAIR data sharing. Developed since 2015 and public since December 2017, it already provides data on materials computed by over 100000 calculations. To our knowledge, currently no other infrastructure exists that is designed to guarantee the same level of reproducibility and FAIR sharing of the results in the field of Computational Materials Science, coupled with automation of simulations and workflows.

The aim of this project is to scale up our Materials Cloud infrastructure to match the needs of the whole scientific community, in Switzerland first, extending then to Europe and beyond, while at the same time making sure that the infrastructure is stable, available at all times, flexible to support even more data types and scientific fields. In fact, while currently focused on Materials Science, the platform and the provenance model are very general and can be applied to many different disciplines, starting from computational physics and computational chemistry.

In the 2021 extension, the efforts will be focused mostly on two components: the AiiDAlab simulation platform, and the Materials Cloud Archive for long-term storage of research data. Our aim is to match the needs of the whole scientific community, in Switzerland and beyond, and specifically targeting simulation laboratories that work in close collaboration with experimental researchers. At the same time, we want to ensure that the whole infrastructure is scalable, secure, and easy to redeploy on the (bare-metal or virtual) hardware of any university or research lab.

Principal investigators (2021)

Steering committee

Principal investigators (2019-2020)