OR/14/061 Challenges encountered
|Watson, C, Baker, G, and Nayembil M. 2014. Open geoscience data models: end of project report. British Geological Survey Internal Report, OR/14/061.|
During the lifetime of the project there was significant disruption caused by the departure of several developers and transfer of the principle investigator role on two separate occasions. The enthusiasm of the remaining team members and the fact the head of Data Science, Garry Baker, kept abreast of developments throughout the project helped to ensure the project remained largely on track.
As briefly mentioned in the previous section arrangements were made in the project planning phase to visit Papua New Guinea and Nigeria but due to events in both countries which resulted in the FCO security risk assessment being raised to High, it was decided by the BGS Senior Management that this visits could not be authorised.
Community creation and engagement
Despite there being significant enthusiasm for sharing knowledge on data models, associated IT architecture and business workflows amongst a wide range of contacts there was no pre-existing community of geoscientific, or more general environmental, data modellers. The aim to create and maintain such a community in such a short time whilst also producing a significant number of outputs has proved to have been very ambitious. One possible reason for the absence of such a community might be that many data modelling practitioners only actively engage in research and development when there is a specific end goal, often in response to the need for a new relational database or because an existing one requires re-engineering.
If an initiative such as the Research Data Alliance (RDA) had existed at the time this project was being planned it may have proved a useful vehicle for such a venture. Another approach could have been to contribute data modelling expertise to multi-national projects that involve the creation of geoscientific databases. The data modelling activities required in the production of databases for publically funded research could and, we would argue, should involve adequate resources to produce publically available data models R2.