I am performing a location allocation problem and want to find the minimum number of facilities that is constrained by a drive time (impedance) and capacity. I have 77 facilities and ~40,000 demand points. The existing option for Minimize Facilities doesn't appear to account for Capacity and the Maximize Capacitated Coverage needs to have the desired number of facilities inputted. So essentially i would like a solution that is a hybrid of the 2 tools: ie I want to know the minimum number of facilities that maximized coverage constrained by capacity and impedance. Hope this makes sense
sort of... from the descriptions given here Location-allocation analysis—Help | ArcGIS for Desktop under the
Location-allocation problem types, none of the others are appropriate mixes either? just trying to get my head around it
Thanks, I went through that page in detail and don’t think what I am describing is possible with the existing tools…at least not in an easy way. I think I may have to iterate through maximize capacitate coverage till I find the answer…for example…input 77 locations to start, run, then input 76 locations, run, et al. until I fin the minimum number of locations that gives the maximium coverage. Going to see if I can maybe write a python script to do this.
There is no way to get capacity added to Minimize Facility (even if you added a capacity onto the facilities, the analysis would ignore it). The workaround, per se, is to iterate through the Maximize Capacitated Coverage until you find the optimal number of facilities, which is what you are already planning on doing. You can automate that process with a python script, run it, for example, until the number of demand points whose status equals Not Reached(5) is 0 (Location-allocation analysis—Help | ArcGIS for Desktop). Hope that helps a bit!
The maximize capacitated coverage requires you to enter the number of facilities to solve for. To find the minimum number, you have to do what you already figured out. If you are going to automate this via a script then do consider using a binary search and not reduce by 1 as it will be more efficient to find the minimum number of facilities needed.
Another thing to keep in mind that a demand location can only be assigned to one facility. It's demand cannot be split between two facilities. So if you have the situation where the demand values are large and no one facility has the capacity to serve it, you could place more demand points at the same location with less demand. So a location with 20 demand units could be replaced by 4 points with 5 demand units each.