There is a research organization for information and systems.

Automated simulation software creates a world map of polymer properties
UMAP plot visualizing the distribution of the polymer backbones. The UMAP plots show the distribution of (a) 15,335 homopolymers in PoLyInfo and (b) 1070 homopolymers calculated in this study. The 21 classes of the polymer backbones are color-coded according to the definition of PoLyInfo. Credit: npj Computational Materials (2022). DOI: 10.1038/s41524-022-00906-4

A research team has published their method to create a comprehensive database ofpolymer properties.

The Institute of Statistical Mathematics in the Research Organization of Information and Science (ROIS) is a co-corresponding author of the paper. Hayashi is affiliated with the University of Tokyo. Machine learning is applied to predict new materials with innovative properties and their fabrication methods. Data is the most important source of information in MI.

Hayashi said that efforts to create a database of polymer properties have fallen short.

Hayashi said that they developed RadonPy to build a database ofpolymer properties. It's the first open-source software that uses atomistic models to account for the behaviors and characteristics of individual constituents, and is the first of its kind.

The program runs calculations to get the equilibrate of the assigned polymers. Specific heat capacities at constant pressure and at constant volume can be calculated once it does. The data can be accessed later if it's produced and stored by RadonPy. Transfer learning is a machine learning technique used to correct biases and variations between simulations and experimental data.

The National Institute for Materials Science's Research and Services Division of Materials and the Institute of Statistical Mathematics in ROIS collaborated on the study.

The program implemented a set of automatic computation functions for 15 different properties which were compared with experimental data to verify the calculation conditions. The agreement between six properties was verified by us.

The research team was able to identify eight high-conductivity polymers. The world's largest open database of polymer physics is being created with the help of RadonPy. Three universities and 19 companies are working together to develop other databases with RadonPy for a variety of applications.

The project will create a world map of material properties. It can't be done solely via experimental approaches requiring a lot of money. The research is the beginning of a new horizon.

There is more information about the RadonPy: automated physical property calculation using all-atom classicalmolecular dynamics simulations.

The research organization has information and systems.