Rooms Optimization

Hi everybody,

I got a problem trying do develop a Room Optimization system.

I'm at the beginning of the script and i'm trying to move different rooms on each perimeter due to have different configurations.

the idea I had is to divide each object into several points and creating a translation vector between couple of points. this works with 2 rectangles, but when the shape becomes more complex it doesen't work any more.

my aim is to move lot of boxes on each other perimeter without generating intersections. any idea to continue or to start again? thank you so much

here my work untill now

Attachments:

2015-09-28 primi passi in avanti.....gh, 17 KB

Replies to This Discussion

Permalink Reply by peter fotiadis on September 30, 2015 at 10:20am

Er ... can we reload the Matrix on that one?

Here's a more "clean" thingy to start talking: now list the rules: for instance what exactly means "optimize"? according to what condition? blah blah.

PS: roll-the-bones demo is just ... er ... a roll-the-bones demo, he he

Attachments:

rooms_optimize.gh, 27 KB

Permalink Reply by Phil on September 30, 2015 at 11:08am

Hi peter!! thank you so much, i've seen your script and is amazing, far away form my abilities :)

but there is a problem, maybe I did't eplained the problem well... the 3 shapes I've done where rooms. I was trying to move the rooms getting different configuration keepeng them in touch.

maybe with an exaplme is easyer. i got a bath, a bedroom and a living room, I want to move all these rooms to make different plan configurations. The rules are that the 3 rooms must have a contac with the others and must not intersect, but them should move free always being in contact.

the optimization will be a further step... I will try to do solar optimization

Permalink Reply by peter fotiadis on September 30, 2015 at 11:48am

OK, got it.

BTW: doing this is one thing (task A) ... but the bad news are that maintaining collections of possible variations made "on-the-fly" (task B) requires code. - and with no other restriction they would be "a few".

I'll try to create the SIMPLEST possible def on that matter (task B), but if you find it too-much ... don't bother: just recycle it, he he.

Here's an analogy: forget rooms et all and focus to some humble 2d grid of points. Now I want from you to access a given point, vary the Z (by means of a slider), store the change go to some random other point do the same etc etc. If you replace Z with the variations of room combos (with varying dimensions) ... you'll get the gist of the whole challenge I do hope.

Permalink Reply by Richard Schaffranek on September 30, 2015 at 11:11am

Hi Phil,

could you explain your rules via a drawing?
Maybe you could habe a look at:

http://www.grasshopper3d.com/video/floorplan-generator

Or

http://www.grasshopper3d.com/video/floor-plan-generation

If those are of interest I can help you with a file for the first case and a paper for the second.

Best Richard

Permalink Reply by Phil on September 30, 2015 at 11:58am

if the gif goes, it rapresent the idea, as you have seen at the beginning of the script, rectangle should change dimensions too. the first case you post is an amazing tool, but I would like to have some flexibility always for the total shape.

thank you so much againg ;)

Attachments:

Senza-titolo-1.gif, 146 KB

Permalink Reply by Kim hauer on October 4, 2015 at 7:51pm

Hi Phil
I don't know what drives designers, to think, a simple rearrangement of some boxes is going to solve, how to design a living unit for human beings. Form has and always will follow function, unless you have a master plan of what each space contains, your design, of shifting boxes is not going to provide you any design solutions. If you look closely you can already see laundry drying on the balcony's of the proposed skyscraper on the right :)

Permalink Reply by Phil on October 7, 2015 at 3:14am

Hi Kim,

Of course a rearrangement of boxes is very far from my final aim. unfotunatelly I don't have already the skills to plan my final idea in Grasshopper, but I'm trying to do that step by step. And boxes that shift is just a first step to create a rule to keep rooms in touch.
Finally, most of all the second example you showed me is as far as possible from my aim, It is an unbeliveable exaple of how to use grasshopper to create really beautiful stuff without thinking at people :)

Permalink Reply by peter fotiadis on October 7, 2015 at 4:01am

In fact ... Orwell type of stuff for numbers (in the near future numbers would replace humans [and humanoids as well]).

Other than that ... er ... hmm ... recently a fellow user (who May be a Robot or worst: a Cyborg) initiated the mother(*) of all AEC threads ... or in fact a small portion of it, he he.

http://www.grasshopper3d.com/forum/topics/adjacency-matrix-to-a-wei...

OK, I admit that this talking is a "bit" freaky (not to mention the C# freaky part)... but read it anyway: in The Name of Science, he he (did I've mentioned that is a bit freaky?).

(*) get rational add irrational, stir serve chilled.

best, I Robot

Permalink Reply by Joseph Oster on October 7, 2015 at 6:18am

When I first saw this question posted, I had the same thought as Kim - do you really think an algorithm of random rooms will ever produce something this beautiful and functional?

Best Small Home 2015 - 550 sq. feet, not including covered porch
https://www.youtube.com/watch?v=RMGGmfwjqxY

http://www.finehomebuilding.com/houseawards/2015/best-small-home

Permalink Reply by Phil on October 7, 2015 at 7:45am

Of course I do. my aim is not to do a random aggregation of rooms. Is step by step, to create a tool to support architects facing hundreds of problems at the same time and trying to find the best compromise between them. that house is beautiful but it could be done in other thousands of ways. Maybe its floor plan is stupid, or could just be done better to optimize natural ventilation, solar radiation o movements of peolple, my aim is to find the best one. I have always to say that it is beautiful for it's interior design. And this is a complete different topic.

Permalink Reply by Joseph Oster on October 7, 2015 at 8:24am

Oh brother... its floor plan is definitely NOT stupid. And while some plans may be better than others in quantifiable ways, "the best one" will prove to be an elusive goal (hence, the annual award for "Best ____"). Tools are great, of course, but architects aren't idiots.

Permalink Reply by Phil on October 7, 2015 at 8:47am

Anyway I have to tell you that I had a skype call with the Autodesk Research Group and they have differt opinions about that. I'm 100% sure that today i don't have the skills to do that, as I'm 100% sure that It is a way to develop architecture and to avoid most of the silly stuff that we see aroud. Of Course there are beautiful houses and good architects, but there is always an idea behind a project and an approach. this is my aim. create a tool that will stop thinking at an house starting from idiot conceps like Chips (Alsop Architects)