My first job was fence painting. Seemingly never-ending work..
Later work, with an architect, is the reason I got started thinking about any of this. The amount of detail we put into modeling was incredible. He and I built them together. Could we make our target, the target of our effort, some other target? I started thinking. But it took awhile.
In 2007 I decided to go back to the old ways again and see what its like to drive every drawing in a construction document set entirely from the model. That’s this project http://www.youtube.com/watch?v=fYUHMT-gTR8
During that project I thought, now wait a minute. These are nice drawings (very nice) and they are completely automated from this model, which is great, but when I give these drawings to the builder he doesn’t know they came from a model. These drawings look exactly like any drawings have ever looked. Their purpose is to communicate, but they’re not doing any better job of communicating than any drawings ever have, and anyway, there are communicative limits to drawings, real limits, because they are abstract (despite their real strengths).
And then, looking at the model, I remembered again that when I hand over the model to the builder, to him the model is not really very useful because it is an overwhelming environment of information. It doesn’t tell you what to look at, or do. And it doesn’t tell you where it is reliable and where it isn’t. So, at one point it occurred to me (seemingly out of nowhere) that those are exactly the things that drawings are designed to do. So that’s an idea:
Drawings exhibit the quality of directiveness.
Directiveness is essential to AEC practice,
and therefore directiveness should be exhibited clearly within modeled environments (BIMs)
I went to work for Bentley in 2008. Now the drawings (directiveness) get aligned into to the model automatically. Now you can deliver the model and it is no longer speechless. It tells someone what to look at, and do, and at the same time it tells them where you say it is reliable. Elsewhere it may or may not be reliable, but the environment of the model makes the directives of the drawings easier to understand correctly and thoroughly. So aligning them together seems like a win for the drawings and a win for the model. Plus everyone can focus their work on the directiveness, instead of trying to model everything.
The same idea is working with point clouds and other visual environments toohttps://dagsljus.wordpress.com/2013/03/07/directive-environments-point-cloud/