Supply chain optimization algorithm - limits per day problem

I am planning to do a PhD in supply chain optimization. I want to solve a warehouse localization problem. I have 120 long, alt addresses in West Europe. I need to calculate distances when using a clustering algorithm, then use a grid to locate the best option for the warehouse. For this purpose I will need to calculate the distances approximately 120 thousand times. I think opensteetmap is the best option for this problem.

  1. Can anybody provide some information from what to start? For now I have an algorithm made with Wolfram Mathematica, However I didn’t find any openstreetmap scripts that would work with this software.

  2. Is there any finished application that I could use and communicate with it easily? E.g. googlemap api works, however it has a 2500 day limit and extra calculations would cost a lot.

3.Is it possible to use simply a web link, that wouldn’t have any limitations? Maybe someone could provide it and for a few days turn off the limits for my IP address?

  1. If no, how should I use openstreetmap to solve this problem? Do I need to download the osm planet file and write the code to return the distance between A and B, or is there any completed solutions? I would need to remake the algorithm with another programming language in this case.

I would be really grateful for any advice or help,

thank you in advanced

See also

Not knowing what the emphasis of your thesis lies on I’d like to question if you need accurate street distances at all. I am doing a lot of these warehouse optimizations myself, for a living, though. Depending on the case I take accurate street distances (provided by some supply chain tool routing algorithm), pre-defined distances matrices or an approximation by bee line. For Europe there is the so called EWS road distance matrix on zip code level. It’s commercial but maybe they offer an academic license and I am not sure how you could use your lat/long pairs. For many cases and especially if your only goal is to proove your optimization algorithms it might be sufficient to use the bee line distance (great-circle distance easily calculated from lat/long) and possibly multiply by some regional factor (e. g. 1.3) to approximate the road distance.