escape route (distance around several corners)

Hi to all grasshopper experts

following problem is what i'm dealing with at the moment:

it is about escape routes in a complex ground-plan.

from each point in the building, anyone should reach one of the exit points in 40 meters.

to simulate this situation (set the exit point, then test if any point is in the radius), i'm always drawing this scheme:

the blue surface is including each point that is reachable in 40m.

Now i'm searching a tool that could help me doing that more efficiently.

I tried with isovist (doesn't work around the corner), i'm thinking about using kangaroo (no idea if it has the possibility to simulate that, because i do not know it enough).

But maybe i'm completely on the wrong way.

Can anyone give me an advice?

Thanks a lot!

Replies to This Discussion

Permalink Reply by Richard Schaffranek on December 17, 2013 at 7:10am

SpiderWeb Sample file...

Attachments:

escapeRouteMesh.gh, 12 KB

Permalink Reply by Vicente Soler on December 17, 2013 at 11:22am

The visibility graph you created is for Manhattan distances (which can be a good idea to use in case the rooms are full of stuff laid out in a grid, like desks in a classroom) but in the previous definitions it was assumed people also walked diagonally.

The main bottleneck still seems to be constructing a proper visibility graph which it doesn't seem that SpiderWeb can help.

In any case, the definition I posted creates a different graph for every start point (since I knew some points were not going to follow some paths) but with SpiderWeb I can just merge them all together to solve it once and it will shave a few seconds which is nice.

Permalink Reply by Richard Schaffranek on December 17, 2013 at 11:40am

Don't know exactly what you are working on but the question popping up is:

Is the extra precision needed?

You can also quantify it as: 2^.5 * length of path. so at a maximum the path calculated is about 41% longer than actually. If you add diagonals between the vertices than it comes down to even less...

Richard

Permalink Reply by Liam Ross on July 10, 2014 at 9:16am

Hi Jonas,

I just posted a file that models escape distance, but also have problems with complex plans. Interested to see what you think of my script, but also whether you got anywhere with the above?

Attachments: