GeoWeb 2009 in Vancouver, July 27th to 31, was one of the most interesting conferences I have been to. GeoWeb is all about the semantic web and geoinformation and this years theme was entiteld “Cityscapes”. So many of the talks in the academic as well as in the business track were about capturing, storing, managing, and utilising 2D and 3D geodata. However, one of the highlights was a keynote by Dr. John Stutz, Tellus Institute.
To watch the missing parts of the video or other geoweb talks visit:GeoWeb Youtube channel.
Something which was obvious is that one of the major challenges right now is to manage all the 2D and 3D data available and put it into context with other geodata and information using web-services. So you would hear a lot abbreviations like KML, WMS, WFS, SOAP, AJAX, CityGML, GML, RFD, SVG, OWL, …
With respect to 3D GIS there have been several talks centered around CityGML. Wiebke Tegtmeier who works at the ICT in the Netherlands Presented a data model for representing Geological and Geotechnical Information which extends CityGML to include natural sub-surface features. Gilles Falquet, University of Geneva, presented an ontology-based approach to combine CityGML-based models with other spatial and functional models. Thomas Kolbe presented an conceptual framework for interpreting graphical 3D models to reconstruct their semantical structure and transform them into CityGML or IFC. Olaf Schroth and I finally presented our e-Collaboration concept which aims at enabling stakeholders involved in planning processes to make better and continuous use of “official 3D city models” maintained by the administration.
I got introduced to Paul Cote who works at the Harvard School of Design. He is the one who collects and organizes all the information and tutorials related to 3D Geospatial Modelling such as how to create a 3D model from Google Earth and SketchUP or how to process TIN models in ArcGIS. He works at creating a 3D city model for Boston from heterogeneuous resources and we talked a lot about workflows. So there might be models created using automatic feature extraction in KML format, architectural models developed in 3DS Max or SketchUp, geospatial models stored as multipatch-features in an ArcGIS geodatabase and so on … The challenge seems to be to find workflows that allow us to read all these models, transform them into one exchange format, add arbitrary semantic information, and store them in a central database. We both are very excited to learn that FME from Safe Software now is said to have a working CityGML Reader and Writer and that they at the same time can load .kml, .vrml, .3ds and in the near future .skp models. So this might open up a path to integrate these heterogeneuous data sources into one database.
Finally, I like to add another Youtube Video showing Javier De La Torres talk about managing Biodiversity data on the web. His web-mapping examples and the databases he introduces are very nice examples of how geospatial services and database services can be combined to handle an huge amount of information.