The Need for Comprehensive 3D City Models (Part 3)

In previous posts on this topic, I have discussed some of the more important communities that have requirements for 3D city models and some of the common use cases that they share.  In this final post in the series, we will look at some of the tool sets that I hope will evolve in the near future.

Required Tool Sets

If we are serious about building comprehensive 3-D city models, then we need to be thoughtful about developing tool sets that are appropriate to the task.  While we are starting to see the evolution of some interesting tools for specific communities, more general and inter-operable tools for the creation, maintenance, publishing, and consumption of 3D city models are still on the drawing boards.  Here is a wish list for capabilities that would be broadly useful across all communities:

Data Models

To begin with, we need to develop data models that are sufficient to support our requirements. There are a number of emerging data models that address various parts of the requirements we have been discussing.  The IFC BIM community has a reasonably mature if extremely detailed data model designed to address problems related to designing and constructing buildings.  While BIM models work well for the problems they were designed to address, they do not scale well and are not appropriate for landscape-level modeling.  The BISDM community data model does a reasonably good job of modeling simple features within buildings. This model supports a pretty reasonable level of granularity, scales and performs very well but does not address fully textured architectural renderings of buildings.  The CityGML model put forth by OGC does a nice job of multilevel representations of buildings, as well as fully textured architectural renderings but does not handle building interiors and requires database specific implementations in order to scale.  In the next few years I think that we will see a maturation of these complimentary data models along with a better understanding of how to harvest subsets of data from one data set for use in another.  One example of this would be harvesting data from a BIM using the space model view for population of BISDM.  The data conversion and interoperability implications of this need for harvesting from one data store in order to populate another are significant.  This will be a guaranteed growth area for the folks at Safe Software.

Data Authoring

Once we define a data model that is appropriate to the scale and scope of our particular city’s requirements and use cases, we will need data authoring tools that are capable of creating and maintaining data in this data model. At current software versions, the process of offering 3-D city model data requires the use of several different authoring tools and the workflows required to move data along in the process are less than elegant.  Lets hope it is not too much longer before the major software vendors provide the capability to easily author fully textured and attributed 3D building models and then exchange that data with other systems.

3D & 4D Geospatial Analysis

For me, the real excitement is the promise of being able to perform true 3D and 4D geospatial analysis.  Again, in present versions of available software, tool sets available to support this kind of analysis are pretty green. While there is progress being made particularly with desktop tools and I have seen some encouraging demonstrations in limited scenarios, we have a long way to go before we can truly deliver compelling geospatial analysis in 4D at a citywide scale.

3D & 4D Web Publishing

In order to truly unlock the potential of comprehensive 3-D city models to support the workflow requirements that we have identified, we will need to be able to consume these models as services published over the web. Given the complexity of some of these models, server performance and network bandwidth constraints become significant challenges almost immediately. That said, making this data available through Web services will be critical to the growth and adoption of this technology going forward.

Visualization on Multiple Platforms

For a solution provider like PenBay, we are very focused on being able to deliver information to our users were it makes the most sense in their daily workflows. In some circumstances, this may be the desktop. At other times a web browser may be the most appropriate delivery mechanism. In many situations, a mobile device is the best way for a user to access and interact with their business data.  The major software vendors will need to build their platforms in such a way that once authored our 3D city models can be consumed by a variety of different devices and platforms.

Data Sharing and Interoperability

I have been around long enough to understand that data sharing is an important and complex issue. For my money, data sharing is much more than publishing KML for simple Google mash ups.  Data sharing to me suggests that the data will be useful for geospatial analysis, that it's metadata will be appropriately documented, and that authentication and entitlement will be thoughtfully addressed. Data sharing and interoperability in the world of comprehensive 3-D data models will likely be complex and challenging. Particularly when we get to the point where different people or organizations are responsible for maintaining the data in different portions of the model, issues of data documentation, currency, and security, become difficult to manage in a systematic way. The very nature of the data itself and the communities of interests that we are serving with 3-D city models requires that we be thoughtful and proactive about designing data sharing and interoperability into our approach from the start.

While various standards bodies have usually taken the lead in establishing interoperability guidelines, this is not the only way to achieve interoperability and to my mind is not necessarily the best way.  Many of the most functional standards like KML, .pdf, and .shp to name a few have been developed by private companies who then published their formats to support interoperability.  There are those that complain that "standards" like the Adobe Acrobat Reader format or the ESRI shape file or file geodatabase are not "open" and are therefore somehow tainted.  From my perspective, I would rather have file formats developed by a team of software engineers who are on the hook to deliver practical, functional, and highly performant software than a committee of individuals who are less accountable for actually making software work.  What really grinds my gears is companies like Autodesk who makes a very popular 3D authoring tool (Revit) but refuses to publish their format so that their data can become more fully interoperable with other systems.  For the foreseeable future, practical interoperability will no doubt continue to be the domain of our friends at Safe Software.

So...  I have talked a lot about my personal vision for 3D city models.  Where do YOU see the future for this concept?  Where do you think the first real practical inroads will be achieved?