Accelerating the Adoption of Research Data Management Strategies

The need for good research data management practices is becoming more recognized as a critical part of research. This is driven by the exponential rate at which data were obtained through high-throughput computations leading to the 5V challenge in Big Data, including volume, variety, velocity, veracity, and value. Poorly managed data often restricts its usability and accessibility, especially when various research groups and organizations collaborate on multidisciplinary research projects. This deters data users because they lack a basic understanding of the data's provenance and conditions of use. The material science community is no exception to the challenges of data deluge as it heralds its new paradigm of data-driven science. This paradigm uses artificial intelligence to accelerate materials discovery but requires massive datasets to perform effectively. Hence, there are efforts to standardize, curate, preserve, and disseminate these data in a way that is Findable, Accessible, Interoperable, and Reusable (FAIR). To understand the current state of data-driven material science, we surveyed researchers working on a small-scale research project and another within a large consortium. We analyze the current research status of each community and the challenges they face regarding the use and management of research data. This enables us to provide relevant recommendations to develop and/or procure an effective research data management system following the FAIR guiding principles. This work positions these recommendations on their urgency within the data-driven research life cycle.