All Things Location

Expand all | Collapse all

What does your GIS stack and processing environment look like?

  • 1.  What does your GIS stack and processing environment look like?

    Moderator
    Posted 05-21-2018 16:51
      |   view attached

    Share how your geospatial data is stored, the tools you use to create and analyze it, and share it with your colleagues or clients. We'll provide a Li360 steel water bottle to everyone who shares.

    See Attachment



  • 2.  RE: What does your GIS stack and processing environment look like?

    Posted 05-22-2018 03:32

    Internally, our data is mostly stored in TAB format and analyzed within MIPro (v16 with a bit of v15). In addition to that, we use SQL Server for storing databases that we use for address fixing (such as fixing city and street names) which is done by our own internal program that uses both synonyms and an algorithm that finds closest matches. We also use some .NET applications that do things such as converting coordinates (we convert Israeli coordinates, ITM, to LatLon and back).

    As for our clients, in addition to the ways we store, many of the large corporations using Oracle in their legacy programs and some moving towards PostGIS now, especially newer departments that deal with big data.



  • 3.  RE: What does your GIS stack and processing environment look like?

    Pitney Bowes
    Posted 05-22-2018 10:15

    Thanks for sharing Kobi!

    Your setup seems similar to that of many clients here in the Nordic market, with a mix of TAB files (for the speed, and being available offline) and spatial databases + also non-spatial for adding related attribute data, such as property ownership.

    A couple of questions come to mind:

    1. Are you also working with Raster/Grid data, and if so, have you seen improvements when working with our new MRR format?
    2. It would also be interesting to here your feedback on the Address quality and normalization provided by PB through the Spectrum platform, if you have had the chance to benchmark this with your own solution built on the SQL Server data.


  • 4.  RE: What does your GIS stack and processing environment look like?

    Posted 05-23-2018 03:10

    Hi @Hakan Karlsson?, my pleasure! I did it for the water bottle :)

    1) No, I don't really use Raster/Grid data all that much. One of our telecom clients has one or two country wide layers for cell quality. As far as I know they use TAB format for that. But they are upgrading a bit now, including moving MXT from 6.8 to 9. So we may have them move to MRR now. But that will take a few months.

    2) I haven't tested out PB addressing. In Israel there is one main provider of vector map data which includes address ranges for the line segments. In addition to that there is a governmental body that sells a layer of house contours with address points. So most client purchase that data which we use as the basis for their geocoding. And the synonyms we use is a DB that we developed internally over the course of some projects (such as 200+ variants of the city Tel Aviv in one clients old database). I may try to give the PB Spectrum addressing a try in the near future and will report back.



  • 5.  RE: What does your GIS stack and processing environment look like?

    Posted 08-30-2018 10:49

    Our setup is very basic. TAB files stored on a server, analyzed with MI Pro v15 and v17. We share files back & forth with our clients as .shp. We do not do much geocoding, as most locations we deal with are rural addresses and we have not found a solution that works well for that. Most of our work has been short projects over the years, but the past year, we have gotten into longer, repetitive work that has allowed us to start learning and using MapBasic to automate many of the smaller tasks. We don't use any spatial data, only Bing Roads & imagery as our base maps.