Speaker
Description
An extensible open-source platform to support digitalization in materials science is proposed. The platform provides a modular framework for flexible web-based implementation of research data management strategies at scales ranging from a single laboratory to international collaborative projects involving multiple organizations.
The platform natively supports object types related to materials science, such as chemical systems and compounds. The extensible types system allows easy introduction of new user-defined object types. To implement deep integration with the types added, an external API is implemented, which is responsible for validation of documents, data extraction from them (for input in the database), and visualization of documents. Like late binding in programming, this allows the system to be extended without changing its source code by delegating the above-mentioned tasks to external web services configured at the object type level.
The system supports configurable templates for table data and properties for user-defined object types, allowing efficient storage and flexible search for material science entities. Build-in reports provide core metrics for data quantification and users/projects contribution evaluation. A special reporting API provides full read-only access to the system with the ability to securely execute arbitrary SQL queries to implement any form of custom reports or arbitrary data extraction for external use.
Multi-tenant support is implemented, enabling rapid deployment of new instances of the system, allowing the system to be provided as a Software-as-a-Service.
The system is here demonstrated as a data repository for combinatorial synthesis and high-throughput characterization data in Materials Discovery and Interfaces group, allowing efficient handling of thin film materials libraries and integration of data from different measurement devices with further flexible search capabilities. Furthermore, its use in the CRC TRR 247 is discussed.
This research was financially supported by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) Project-ID 388390466-TRR 247 (subproject INF).