Proposing a digital data steward as a Service for project-oriented research

As the first CoRDI showed, many solutions to RDM problems are already available. These reach from holistic, conceptual models like data life cycles to specialised tools for specific RDM tasks. Many of them focus on project-oriented research [1], [2], [3], [4], [5]. Such project-oriented research is prevalent in all types of research across the spectrum from basic research to applied research [6]. Especially in engineering sciences, project-oriented research is a dominating way of conducting research for a long time [7], [8] and is fostered by project-oriented allocation of funds by funding organisations [9], [10]. As such, RDM tools directed towards project-oriented research are a direct result of the structure of the research landscape. The next step in the development of RDM is to consolidate existing solutions and agree on common best practices. Especially currently, with the emergence of new technologies originating from the field of machine learning and artificial intelligence, well managed data is crucial [11], [12]. In this context RDM serves as a way to address rising challenges like globalisation, digitalisation and the need for increased efficiency [5], [9]. Researchers often miss guidance in RDM, especially in project-oriented research [13], [14], [15]. The integration of RDM into everyday workflows is hindered by the effort required to initialise and conduct RDM [13], which in itself is aggravated by needed enquiries do figure out how to conduct RDM. These enquiries are dependant on the circumstances of the research conducted. For example, RDM can be dependant on policies of funding organisations [16], [17], [18], [19], [20], [21], or universities [22], [23], [24]. Despite their influence on RDM, these RDM policies are perceived as insufficient by researchers [13]. Hence, researchers have problems implementing RDM into their research. The Joint Assistant for Research in Versatile Engineering Sciences (Jarves) strives to address these problems by providing a process for RDM [5]. By ordering the process steps by their occurrence in the engineering research process, Jarves provides aHamann et al. | CoRDI structure for RDM in engineering. Within the process, researchers can select their specific circumstances. Based on these circumstances, the researchers get information on how RDMactivities should be conducted to fit the corresponding policies. This encompasses, for example, available tools and suitable trainings. Partial automation further reduces the effort needed in RDM, as it allows for seamless (meta)data exchange. While the tool originates from engineering sciences, researchers in other domains can apply its concepts and possibly transfer them to their needs. Hence, this contribution to the second CoRDI aims to present Jarves' idea and underlying architecture. Furthermore, a Jarves' current functionalities are demonstrated. If the contribution can be conducted in a workshop format, feedback on Jarves will be collected on how to make Jarves a overarching service for the NFDI4ING and eventually NFDI communities. Additionally, the possibilities for its integration it into Base4NFDI and the EOSC service portfolio will be discussed with the participants.