GeoWeb 2009 Trip Report

05 August 2009, 11:24am

Held annually for several years now, the GeoWeb conference in Vancouver, Canada is not just another marathon of obscure technical presentations or a round-up of the latest product offerings from marketing departments. Rather, it serves as a forum for new ideas and technologies that help shape the discussion of what it means to bring geospatial support into the Internet (and vice versa), and the July 2009 conference was no different.

Photo: The conference centre in Vancouver

Report by Michael Gerlek

The week began with two days of workshops. I had the privilege of working with Ron Lake (founder and CEO of Galdos and chair of conference) and Carl Reed (CTO of the OGC) to teach an all-day course ambitiously entitled “GeoWeb 101”. We introduced the attendees to all the key technologies underlying the geoweb, ranging from location support in the IPv6 stack to the idea of object modeling with GML to examples of large SDIs (spatial data infrastructures) in use today. The other 13 workshops included hands-on experiences with Flex, in-depth lectures on GIS servers and services, and a walk-through of the open source GIS tool stack.

The program Wednesday through Friday consisted of 50 forty-minute presentations, two panel sessions, and six invited speeches. This year’s conference had a special focus on “Cityscapes” – the integration of architecture, urban planning, and BIM (Building Infrastructure Modeling) with the geoweb itself. In support of this theme, the conference organizers invited a number of speakers who had what might be considered nontraditional topics and backgrounds.

One of the invited speakers, for example, was Ken Greenberg, a well-known architect and urban planner. Greenberg made a very compelling case for designing cities with a much more cross-functional and behavioral view than has been practiced for the past several decades. In two projects he termed as “reclaiming land left by the retreat of the industrial glacier” – the redesign of Toronto’s waterfront and the development of post-Big Dig Boston – he showed slide after slide of how he and his team have been able to simultaneously consider ecological, economic, transportation, and residential/commercial needs. As GIS practitioners, we have a large role to play in supporting the complex models and visualizations that 21st century urban planners will require.

Other invited speakers included Alex Miller (President of ESRI Canada), Michael T. Jones (CTO of Google), and John Stutz (founder of the Tellus Institute). In a mainstream way, Miller discussed how geospatial technology can help inform the discussions of the challenges we face today, including climate change and transparency in government. Jones, who has keynoted this conference a number of times, talked about the pace of development in our industry, noting that YouTube didn’t exist 5 years ago and yet is today streaming billions of videos per month. He also discussed how simply having lots of data is not sufficient, using the example that a phone book not sorted alphabetically is of little utility. Most compelling, however, was Stutz’ talk on “The Great Transition”: he argued that the pace of human development proceeds logarithmically, not linearly, and – underpinning Jones’ remarks – we as a society are at a key (and steep) inflection point today. Again, GIS professionals can contribute to our understanding of these societal issues.

The first panel session was on API models for web services: REST versus SOA versus event-driven/P2P architectures. All three models were well represented, but the audience favorite was clearly REST because of its simplicity and correspondingly low barrier to entry. The second panel was on business models for location based services. Here in 2009, it is clear to me that the real and interesting issues for doing business on the geoweb have little to do with making money via geo-targeted advertising on social networks; rather, businesses are looking to augment their existing analytical tools and models by exploiting new and existing geospatial data and data services.

The technical presentations covered a broader, if more typical, range of topics. Clemens Portele (Interactive Instruments) looked at the “gap” between what SDIs are providing today and what we could get with a more geoweb-based perspective. For example, too many SDIs are designed for users who already have domain knowledge; one could envision the geoweb as providing a means for better understanding what data is available from the SDI, what it means, and how it relates to other data. Xavier Lopez (Oracle) described the growing need for 3D support and interoperability in our tools, specifically databases. This is being driven by the increasing use of lidar data sources and CAD models. With the usual compelling demos, Michael Ashbridge (Google) introduced three new extensions to KML 2.2: ocean modeling, scripted playback (“touring”), and support for time series.

As has been noted of a lot of conferences lately, attendance was down this year compared to 2008. However, the variety and caliber of the presentations and the opportunities for networking with some thought-leaders in our industry more than made up for this. The GeoWeb conference continues to provide technological insight into future directions without ignoring the societal impact of our work.

Michael P. Gerlek (mpg@lizardtech.com) is the Director of Engineering for LizardTech. He was a member of the program committee for GeoWeb 2009.

GeoWeb 2009

GeoWeb 2009 presentations on YouTube

Michael P. Gerlek

Director of Engineering, LizardTech.