Currently, there are metadata schemas (such as DataCite, or DDI) that can describe the concrete outputs of research, e.g., papers and datasets, but relatively little work has been done on finding a metadata schema for the research itself. The problem is that different disciplines have vastly different ways of organising research activities, for instance because of differences in funding models and mechanisms, or in requirements for approval, and thus differences in how and when research is split into discrete activities and labelled.
You can read the full story on Zenodo here.