Rapid Establishment and High Returns
of Material Genomic Data Platforms
Wang Zhuo1,2*
1 Chengdu MatAi Technology CO., Ltd, Chengdu, 610041,
China
2 Light Alloy Research
Institute, Central South University, Changsha, 410083, China
ABSTRACT: As a new approach to materials science within the fourth paradigm of
science, materials research and development requires tools to enable it. A
materials genomic data platform (MGEP) software will integrate materials data
collection, data management, software integration, and comprehensive data
applications. Lack of data is a bottleneck for global materials research. For
the collection of experimental equipment data, literature data, and material
calculation data, the MGEP provides tools to realize rapid and accurate
acquisition and analysis of various types of data. Especially for literature
data collection, MatAi has successfully developed a literature data processing
system, which realizes the convergence and integration of high-value data in
massive literature through digital analysis, extraction and structured storage
of literature data. Material data is massive, complex, dispersed, fragmented,
and constantly growing, the MGEP needs to provide systematic and structural
design functions with good scalability and flexibility, and construct high
quality and standardized gene data set for the whole life cycle of materials.
Based on Integrated computational materials engineering (ICME) and Materials
Genome Initiative (MGI), the MGEP integrates cross-scale material computing
software from quantum mechanics to material service, and invokes
high-performance computing clusters based on concurrent computing driver
engines to achieve integrated management of material computing simulation and
data integration. The integration of machine learning, artificial intelligence
and other technologies with the MGEP will provide new ideas for understanding
the relationship between material composition, process, structure and
properties. The MGEP will effectively mine the patterns hidden in material
data, improve data utilization, shorten the material development cycle, and
reduce material development costs. In this paper, the establishment and
application status of MGEP under the framework of material genetic engineering
will be discussed.
Keywords: data-driven;
data platform; data acquisition; data management; machine learning;
Wang Zhuo completed his master degree in Information and Computational Science from Northeastern University (China). He is the CEO of Chengdu MatAi Technology CO., Ltd. He has long been involved in the research and development of materials databases, computational materials science, and the construction of materials knowledge systems.