Large-scale cyberinfrastructure projects are suggested to be the solution to many of our current scientific research needs. Within the earth sciences there are many different groups working on these types of projects. Examples can be seen in the various NSF supported projects under EarthCube (i.e. iSamples, GeoLink, GeoSoft). While these projects are still in development, international communities such as the Research Data Alliance (RDA) and the Earth Science Information Partners (ESIP), are pushing to ensure that these various project communities talk to each other. This will ensure their work is interoperable and that they do not duplicate efforts already undertaken in different domains.

Once complete these projects will be a package of software, practices (both scientific and mechanical), and conceptualizations for conducting scientific research. Once taken up by the community, they will be the ‘standard’ way of completing research – supported by peer reviewers as the authoritative method for conducting a specific kind of research. We can consider these large-scale cyberinfrastructure projects to be in essence, standards. An important part of the standardization process is that ‘standards’ must be taken up by the ‘market place’ in order to be successful (Cargill, 2011).

The individuals, the domain scientists, the researcher in a lab or the field – they are the market place. As cyberinfrastructure is introduced, we must ask those in the market place to adapt. This includes developing as well as training in the new sets of skills needed to interact with these standards. These skills include proper data management – both for new and existing research data. This talk will address the limitations inherent in these large-scale systems, and more importantly the future of these systems: 1) what changes are needed to include these 'standards' in our current process of training the individual scientist? and, 2) how as a community of individuals, can we make these changes happen?

Cargill, C. F. (2011). Why standardization efforts fail. Journal of Electronic Publishing, 14(1).