By Randall S. Newton
Let’s talk Building Information Modeling (BIM) for a bit. I love BIM The Concept; it’s BIM The Implementation I can’t stand.
I love BIM The Concept. Putting all project information in a central repository, accessible by all team members, is a grand idea. Make the geometry 3D and keep the underlying data open and sharable. BIM The Concept is about re-engineering work processes and collaboration through the disciplines.
But BIM The Implementation leaves a lot to be desired. In search of construction data nirvana, we go to the CAD store, only to find Eeny Meeny Miney Moe as a software deployment strategy. With apologies to the fable from India, BIM has become the AEC equivalent of the elephant, and we are the blind men groping it for understanding. Is BIM a cat’s cradle of linked DWG files (Autodesk Architectural Desktop)? Is BIM a front-loading bucket for a much larger proprietary approach to construction engineering (Bentley Architecture, or whatever they are calling it this quarter). Is BIM a virtual model of the architectural geometry? If so, is it based on open (Graphisoft ArchiCAD) or proprietary (Autodesk REVIT) data storage/retrieval? Is BIM a new wrapper on an old solution (Nemetschek ALLPLAN)?
The problem isn’t so much about the specific software product, as it is about the need to make a decision at all. Once a firm standardizes on Revit or Bentley Architecture or VectorWorks, they are pretty much stuck in a narrow universe created by the vendor. As Lawrence Lessig tells us, in cyberspace code is the law. Protestations to the contrary, real interoperability is still a dream; the CAD vendors are the law, and when we use their tools we are limiting ourselves to the universe of their design. Building information modeling should be about data transparency—the ability to move live, intelligent data from one application to another, from one stage of the project to another. BIM should be about more than geometry and design; it should be about engineering loads, punch cards, and construction status. The model should be able to account for time as well as space, progress as well as design. And before we stamp BIM as done, let’s make sure the Facility Management guys have something to say about the dataset.
Here’s the BIM I want (call it BEM—Better Elephant Model). The GUI is SketchUp from @Last. The data is stored in Oracle (or mySQL, if you prefer an open source option). There are hooks into Microsoft Share Point, and the system understands Perl. The geometry engine is from Rhino; the 3D feature set from SolidWorks. You can look at the project as geometry and data side-by-side, assuming your firm doesn’t cheapskate its way out of buying multiple monitors for professional staff. Google support is built in. Parametric? Oui, mon ami. The whole thing is IFC compliant and it even works as a Groove Virtual Office application, as will the modules for the rest of the AEC and owner/operator team. After all, this is just the design geometry engine. There’s still structural analysis, construction planning and sequencing, energy analysis, estimating and job costing, facilities planning, asset management, and many other functions to cover.
Let me leave you with one bit of information that says more about the state of BIM for AEC today than all my musings. Recently the global structural steel and façade specialists Permasteelisa decided to upgrade their CAD from 2D AutoCAD to something that could do 3D modeling. After evaluating all the choices on the market, they went with none of the available AEC offerings. Instead, they chose SolidWorks, ostensibly a mechanical CAD product, to do their structural design work. And they bought 1,200 seats.
