I found a project by Joel Simon called Evolving Floorplans (via Kottke) which is all about the algorithmic generation of floorpans, specifically in reference to a school design. Evolving Floorpans is the first time I have seen a clear visualisation of an algorithmic architectural design, it’s quite beautiful.
Firstly I should note I’ve no experience in the writing of code such as for the alogorythm that Joel wrote and only an interest in the experimental zone of 3d architecture printing, so having said all that some observations;
Just programming for efficiency of circulation gave a remarkably ‘organic’ plan structure but maybe this is not be surprising given how things are designed ‘bottom up’ in the natural world.
The plans, optimised for traffic flow didn’t translate at all for classroom use, building usability or internal legibility. Square up the classroom sizes a bit and use something that errs towards more modern classroom layouts like joined classrooms etc and lab spaces. The structure was not addressed, and we probably shouldn’t kid ourselves that new structural materials or techniques will either come soon or deliver greatly enhanced results than right now. The hierarchy of the room types could also be factored in, ie. the gym and cafe of larger volume, environmental site factors accounted for. Joel already understands much of this though,
‘The metrics could be expanded to include terrain maps, sun paths, existing trees and other environmental input, allowing the buildings to be highly adaptive to their context. The physics simulation could force certain boundary shape constraints.’
If developed further this could easily inform outline design and progress until some baselines and agreements in the basic design have been set. It could provide very soon a tool to help the design team and Architects, how much further it could go I don’t know.