An interesting event that helped to crystallise my views of the level of detail we should engage in when modelling relevant University processes. There were several presentations from JISC projects that had engaged with modelling processes using the ArchiMate language though using different tools. There was also a brief introduction to the JISC Innovation Base which resulted in a partially complete contribution of our proposed modelling activities for inclusion in the emerging IB.
A couple of presentations were from completed JISC Enterprise Architecture projects where substantial parts of university people, processes and systems are being modelled. It was clear from these that there is a considerable amount of time and effort required to both capture and model University wide activities, though there are potentially considerable benefits for the institution. In both cases there were identifiable full time staff involved in the modelling activity and such staff are based outside the typical University structure i.e. not associated with any particular strand of the University. They could be seen as the guardians of the models of existing organisational roles, processes and systems. Both groups were using a tool called BizzDesign which includes support for ArchiMate – this is an expensive modelling tool [ £2000 per year per licence was quoted].
A third presentation (by Alan Paul from APS Consultancy) showed part of the modelling done as part of the MMU SRC project. This was at a lower level of detail and used MS Visio but it was more detailed than the modelling that I have done for PC3 in that roles and data are separated out from the processes and rules used to represent our Academic Award Validation Process.
With regard to the overall workshop there seemed to be an implicit expectation that participants were already using ArchiMate as there was no actual introduction to the language semantics, similarly there were references to approaches such as ‘The Open Group Architecture Framework – TOGAF’ and an expectation or awareness of the levels of support offered for the language within various tools. It was clear during the discussions that there are shortcomings with regard to ArchiMate in that it appears to be well suited to modelling (and visualising) the relationships between roles, processes and systems but was less suited to more detailed process descriptions particularly where there decision points with alternate paths that could be followed.
I propose that we look at my existing models of the Academic Award Validation process and represent it with the roles and data elements split off and then present it back to SRO to see if they find it more useful.