JH: As a dominating theme within research and infrastructure, “Collaboration” will gain increasing importance.  The challenge is how to form research foci within faculty and across research teams that can collaborate?  The international approach seems very organic while the government organised approach can be quite artificial.  In general, it is going to become harder to collaborate at the existing “email” level of communication because of ever greater specialization and an increasing need for collaboration to be real time.

SH: Cross-discipline collaboration is getting harder as well, because eResearch is “tool driven” – and different disciplines have very different tools.  These tools enforce different languages and ways of handling and thinking about data at the local level where they are used.  Therefore, the tools imposed limitations on collaboration between disciplines and also beyond local research.

JH: The risk is the local eResearch analysis and data tools in each discipline will reinforce the silos across disciplines.

JH: Another key theme for eResearch and adjacent to collaboration is Metadata – how do you move your data from one discipline to another?  Many datasets are still very “personal” to the research who created them and often quite specific to that researcher’s needs.

SH: The changes need to be in publishing of data and a move the individual to the institutional level for data management and then to cloud services for collaboration.

JH & SH: Data publishing in NZ is still very organic and self-driven by researchers.  There are few standards or accepted tools at this time.

JH: Noted the move at Lincoln University to an “open data” policy, and also noted that the government is positive about “open data” – but that there is slow progress from policy to implementation to date.

SH: Privacy is also still a key constraint in “open data” and sharing datasets, especially in social and health disciplines.

JH & SH: While datasets can be anonymised; offering access to multiple, cross-referenced, anonymised datasets makes the anonymity far less effective and reduces the strength of privacy protections.

JH: What’s required are “Domain Specific Modelling Frameworks” for data from multiple sources over time, such as exists for land information in Brisbane.

JH: While not specific to the eResearch domain, Visualisation will be a major element of effective research and eResearch in 2020.  How we work with data at all levels will need to change to enable data-driven science to progress in every discipline.  In 2020, I expect to be able to intensify and manipulate a model of weather system with a wave of my hand as easily as I can zoom on an excel spreadsheet today.

JH & SH: to achieve some of this outcomes change is required at a multi-institutional level.

Question: what are the barriers to change in eResearch field?

JH: The IT staff on campus, while they’ve got a job to do, are usually the barrier to be overcome when you try to change something for eResearch.

SH: There is also a lot of stress between the cultural norms in academia, the pressure to innovate, and the need to be efficient.  This results in a high degree of perceived “institutional inertia” when advocating for change.