Information Storage and Retrieval System for a Molecular Dynamics Based Digital Twin of Materials

Digital twin technology becomes an appropriate respond to the new challenge of rapid development in science and industry. Chemical data include the information about structure and properties of materials and compounds. Machine learning technology can help to predict properties of innovative compounds and material, but it takes enormous amount of raw data. The implementation of digital twin technology in spectral and molecular dynamics helps to reduce risks introducing the innovations. Large systems technologies are used in multiscale modeling of materials with computer programs. The world experience of chemical data aggregation and storage is based on specialized databases creation. Disunity of data sources demands a single multiscale data model to operate raw data and amend the accuracy of prediction. To collect and analyze unstructured data full-text search and distributed computing were applied. In respond to perpetual raw data expansion a Data Lake technology is used to store, analyze and search for information of atomic and molecular systems.