anyplaceucy
commented
3 years ago This is one of the long standing issues we have to address at some point (postgis is the perfect candidate). Having a native GIS layer will bring geometry into Anyplace that is currently missing. Any community commitment to this is appreciated.
https://www.cs.ucy.ac.cy/~dzeina/papers/sigspatial17-letter.pdf 4.2 Modeling Directions The technology road-map is towards indoor GIS integration where indoorgml.net, geojson.org or any other standard that may appear in the future and become fully inter-operable. Having the right modeling primitives will give rise to a variety of data management and query processing challenges in the future. Another direction is towards automated indoor model checkers (e.g., graph connectivity, automatically connect overlapping stairs and elevators) as these will make the mapping of an indoor space more straightforward. There already competing industrial systems, like mazemap.com and micello.com, which work directly with Autodesk’s Industry Foundation Classes (IFC) data model that is used for the description of building and construction data. One final direction is the provisioning of libraries for managing specific types of objects in indoor spaces (e.g., office equipment, industrial appliances) but also libraries for representing indoor spaces more richly either using 3D models (e.g., indoor.io) or augmented reality.
alibama
i know this is a big question, but would it ever make sense to build this off of a GIS standard like postgres? opens up some doors and gives a hard schema and simpler apis, still json...