Archinect
anchor

What are the protocols or methods for testing archituctural designs??

Michael Spark

hi,

i want to know if there is a procedure for testing the designs clarity ,

for example, in airports and after finishing the civil work, a test is arranged by asking a test group to go to different gates,and see if any passengers get lost or can't find their gate.

this tests how well the gate numbers and other direction signs are arranged and how visible they are to people.

i'm a software engineer, and i'm developing a system to test this in early stages (before the civil work starts ), so i want to know if there are standard procedures or protocols for this, any information, links, documentation would be very helpful.

 
Nov 14, 14 9:40 am

Nope, all we have are a bunch of critics who wish they were architects. 

Keep in mind that some designs are intentionally obtuse, like self-check in at airports, which is clearly designed to make you want to buy an upgrade. 

Nov 14, 14 10:27 am  · 
 · 
x intern

This is a strange question...  You are writing a program to analyze drawings for way finding clarity?  I don't know that computers can accurately predict human behavior but I'm not a software engineer so maybe they can.  The only way I could think to test this would be to build a 3D model and use it like a 1st person video game using real people to try and navigate the space.  Interesting in theory probably expensive in practice.  

Most buildings if built by an experienced designer of the building type are based on the precedent of the building type, meaning a user will have some experience with the type of building and expect things to work in a predicable way.   If you vary from this significantly it will confuse people and then you need signage everywhere.  

Here is a general example.  When you approach a building typically the entrance is clearly indicated through design elements, if this isn't clear people will end up circling the building looking for a way in.  Once in the building people expect to see a space that indicates where they should go to get to their destination.  If you just dropped them into an unmarked corridor they would just aimlessly wonder around and probably go into the first open door they find.  Each of these is typically handled in a predictable way that people can intuitively figure out based on the type of building.  

Nov 14, 14 11:19 am  · 
 · 
chigurh

It sounds like the generic algorithm used to grade the graphic portions of the ARE...It just pushes around various size circles to make sure there is enough clear for code compliance...

How do you account for the IQ of the user in your proposed software?  How to you quantify the effectiveness of signage?  or designed/implied navigation as suggested by robmc?  It is more touchy feely than coming up with a use value on a scale of 1-10.  Good luck.  

Nov 14, 14 11:40 am  · 
 · 
Wilma Buttfit

You want to write a computer program to test if a complex, multi-dimensional, multi-sensory cultural artifact "works"? Why not just use your brain? Brains are still a million+ times smarter and more efficient than computers when it comes to stuff like that. Interesting idea though. 

Nov 14, 14 11:52 am  · 
 · 
proto

it's a complex notion...like recreating a human in all its interactiveness and independence. the success is determined by millions of criteria, not just one or two or even ten

 

it would certainly be an accomplishment if you were able to do it

Nov 14, 14 12:09 pm  · 
 · 
gwharton

There have been some efforts at creating performance-based design criteria, but none of them have shown much utility outside of narrow ranges of applicability. Performance criteria usually require linear analysis and strong quantification in order to have any rigor at all. Whether or not an architectural design "works" is so highly dependent on interrelated and interdependent value judgments that it's nearly impossible to analyse with linear methods and actually impossible to quantify. The best you can do with quantification is use proxies, a method which is conjectural at best.

Having said that, there has been some interesting work done in this area. Performance criteria that focus on engineering issues (energy, structure, etc.) tends to yield informative results. Design quality, however, has so far been mostly resistant to these kinds of analysis. The only promising work I've seen done on this so far tends to focus on the neuropsychology of aesthetic responses. Arthur Stamps has written up a number of fascinating studies he's done on how people judge design quality in design review.

Nov 14, 14 1:21 pm  · 
 · 
awaiting_deletion

Architectural Graphic Standards (the heavy book)

Nov 14, 14 6:39 pm  · 
 · 
Michael Spark

thx, everyone for your time.

it was helpful.

Nov 27, 14 11:45 am  · 
 · 

Block this user


Are you sure you want to block this user and hide all related comments throughout the site?

Archinect


This is your first comment on Archinect. Your comment will be visible once approved.

  • ×Search in: