location allocation?

166
2
05-26-2011 11:57 AM
Highlighted
New Contributor
I am new to network analyst and I am trying to solve the following problem: I need to find optimal distribution and minimal number of facilities needed to cover 80 % of population of an area with everyone having access to the nearest facility within 10 minutes (that would probably be the specified impedance cutoff). I don't have any candidate facility specified, only the network of roads and distribution of population. Is this possible with location allocation analysis layer, with minimize facilities problem type? Or are there any other possibilities to solve this within Arcgis 10?

I would appreciate your help very much.

Thank you.
Tags (2)
Reply
0 Kudos
2 Replies
Highlighted
Esri Regular Contributor
You can use the Target Market Share option of Location-Allocation to solve this problem. Set the Impedance cutoff to 10 minutes, Use the default Linear Impedance Transformation and in the Target market share set it 80 percent.

Now load in your demand locations as demand and make sure the weigth is 1 (default). That is each demand is equally weighted and thus 80 percent market share will mean 80 percent of the demand locations.

You need to load in candidate facilities. If you do not have any, you will need to "manufacture" some. For example you could load a random sample of the junctions. Or use the Create Random Points tool and constrain it with the street line feature class. You can specify some distance they need to be apart, say 200 meters and you could generate a couple of hundred points to serve as your candidates.

Solve and you will have the minimum set of facilities that will be needed to serve 80 percent of the demand in 10 minutes.
Note that if some demand locations are in disconnected portions of the network (due to bad data or restrictions) you may get facilities located next to them just to serve them. So do anaylyze the results and adjust the inputs as necessary.

Jay Sandhu
Reply
0 Kudos
Highlighted
New Contributor
Thank you very very much. Your advise was really helpful.
Tana
Reply
0 Kudos