As more of my work involves now the day to day use of a common computing model - I thought readers would find it interesting to compare the concept to the Building industry who already have a similar process of their own. Its called BIM - Building Information Modeling. BIM involves creating essentially a 3D replica but embedding within it all the relevant building information including the project plan. In other words its cross referenceable. It is ironic that the Building Industry are using computers (essentially Autocad and similar) to build models - and that my research and recommendations are for the IT industry to adopt more formal modelling! As a further note it is important to realize how complex large scale buildings have become and the risks associated with them. In my view large computer systems are riskier still because they are very complex and less tangible in their construction.

I interviewed Charles Eastman one of the foremost researchers and academics in this space to get a better sense of the approach - one of the key messages he gave me was that the typical sequential A design B plan and C cost paradigm was changing. With earlier visualization and dependency modelling (what happens if we move this wall?) allows design vs cost and feasibility trade-offs to be addressed in an almost parallel mode up front. Implications for IT projects are profound here and I will pick up this strand in subsequent discussion threads.

I like this quote from http://bim.arch.gatech.edu/?id=402 which explains the concept quite well -
“BIM involves representing a design as objects - generic or product-specific, solid shapes or void-space oriented (like the shape of a room), that carry their geometry and attributes. The geometry may be 2D or 3D. The objects may be abstract and conceptual or construction detailed. Composed together these objects define a building model”.

The point that I raise with clients and as part of the ongoing research is that we should work more consistently with “object neutral” computing models - whether simple or complex - to address issues of complexity and mental model differences between collaborating teams. By creating a general category of computing objects and using it as a basic Common Computing Framework - this becomes a simpler and repeatable process. More on how this is done and some tips on best practices in my next post.