Materials scientists and chemists will need machine learning tools to enhance their R&D, according to an IDTechEx report, “Materials Informatics 2020-2030”.
The report said that integrating these underlying operations will not happen quickly, but overlooking the developments in materials informatics will lead to a loss of any competitive advantage.
The report said that Materials informatics (MI) apply data-centric approaches to materials science and certain chemistry R&D. This will be a common method in a research scientist toolkit, and rather than grabbing the headlines, some form of MI techniques will be assumed in all developments.
The key to MI is around the integration, implementation, and manipulation of data infrastructures as well as machine learning approaches designed for chemical and materials datasets, which is why this is an area waiting for channel sales.
Machine learning itself can be used in multiple different projects from finding new structure-property relationships, proposing new candidates or process conditions, reducing the number of expensive and time-consuming computer simulations, and more.
ML approaches can take numerous forms of supervised and unsupervised learning methods; generative models can be effective at screening for optimised outputs across organic compounds while even simple modified random forest models can be useful for proposing follow-on reactions to meet a desired set of criteria. This is still at an early stage with a lot more development required; much can be leveraged from existing developments in AI but integrating specialist domain knowledge and coping with the unique challenges of a materials dataset is essential, the IDTechEx report said.
The application space is broad, and studies have shown success ranging from organometallics, thermoelectrics, nanomaterials, and ceramics through to many more.
