Creating self-driving materials modelling laboratory using AiiDA

In search for alternative and sustainable solutions to achieve net zero targets, design and discovery of novel materials are crucial in a wide variety of applications such as electrification of transport. This process requires exploring a practically infinite number of materials. In this regard, computational techniques have been proven to be a vital tool in making such searches possible and accelerated. However, the large number of materials and numerous calculations which are needed, makes managing those computations challenging and also creates reproducibility issues as such preventing us from efficient utilisation of computing resources.

To overcome such issues, AiiDA (Automated Interactive Infrastructure and Database for Computational Science) provides an excellent open-source framework that provides a modular interface to the computational tools. Hence, a wide range of computational tasks which rely on different calculation types/codes can efficiently be orchestrated via AiiDA-based high-level scientific workflows. In this talk, AiiDA, its underlying structure and capabilities will be introduced along with various open-source plugins and workflows which are ready to use with codes such as VASP and CASTEP to explore materials and their properties.