We urgently need ever more highly optimised materials to overcome current economic and societal challenges, in particular decarbonising the economy. A number of proposed net-zero CO2 technologies are expected to be introduced in the next 5 years but some of the materials they require are not yet at our disposal and traditional materials development cycles can be as long as 20 years. Given the time constraints, new approaches to material development are essential. Data-centric methods of materials design will allow us to maximise the insight we extract from the fruits of expensive research, both by providing new routes to understanding and by ensuring that traditional methods can draw upon the largest possible repository of information.

A key barrier to greater exploitation of data-centric methods of materials development is the lack of robust databases of significant scale. Collectively, we generate new materials data at a colossal rate, and at significant cost. To be useful, this data must be available for interrogation. The more available it is, the more information may be extracted from it. Unfortunately, by default, scientific data tends only to be available to its immediate originators. As scientists, we often produce our data to answer only certain narrowly defined questions and so we tend only to extract a small fraction of the information from our expensively obtained results before they are lost to posterity. Much of the value in our data is never realised.

To enable new ways of working we need to ensure that data is kept in a suitable form, safely stored and referenced from an open database which can be searched on the basis of metadata, which fully describes the nature of the data and how it was acquired. And someone must take responsibility for the curation of this database. There are several key pieces to successfully realizing these changes: we need the infrastructure (both hardware and software) that will store both research data and the databases that index them; we need clear, well-designed processes by which data can be stored and registered, along with suitable metadata; we need to encourage and support the changes in behaviour that will be required within our community to adopt best practice.

This programme of research will:

- Establish the repository infrastructure required to begin storing data generated within the Royce Institute.
- Develop and deploy software for the indexing of materials data and link it to the data storage repository.
- Employ data curation specialists to facilitate the introduction of the storage and indexing systems and to advocate for the adoption of new modes of practice in the research community.
- Ensure that the databases and software tools that we develop are open to exploitation by the whole materials science community.
- Support pilot projects which will pump-prime the new infrastructure with a critical mass of data and develop the new tools and practices to make use of it.

The project will consist of two strands. The first work package (WP) will put in place the hardware and software infrastructure required to properly manage and exploit our materials data and provide dedicated curation resource to help establish its use. The second will fund a broad range of community-led pilot projects that will establish the use of the infrastructure in WP1, develop new tools and processes for exploitation of materials data or establish proof of concept for data-centric methods.