I spent an interesting day in Stratford ahead of the AGI conference this week, at an Oracle Spatial special-interest group organised by the Oracle User Group. Oracle Spatial is the mapping and spatial analysis add-on to the main platform from the database giant.
Oracle occupies an interesting position in the GI world: at once a significant challenge to established GIS vendors, and also challenged themselves by online mapping and data platforms. Would Release 2 of Oracle 11g make clear how they will move forward, I wondered?
The new spatial features in Oracle 11gR2 are certainly impressive. New functionality includes more complex network analysis including hierarchical shortest path analysis and a travelling salesman algorithm. It all felt good to me, perhaps because it makes the database technology seem more, well, GIS-like.
Enhancing usability
Speakers touched on some intriguing ways Oracle databases are powering applications with enhanced usability. Olivier Bucaille from Autodesk advocated using wizards and preset analytical environments to increase accessibility for every user, which I wholeheartedly agree with, and also 3D mapping and graphics at the building scale, which I’m less sure I support.
(I’m not a visualisation Luddite, but 3D bar charts have questionable legibility just used in a document – overlaid on a 3D map they can ask even more of the reader. In addition to this, there are significant statistical issues with taking aggregate data and assigning it to back down to the scale individual buildings in 2D or 3D, pretty as they may look.)
Evolution or revolution?
Anirban Acharya from Infotech Enterprises predicted that use of Oracle databases would increasingly become a key differentiator, and signal the end of proprietary database engines supplied by the main GIS suppliers. His suggestion that spatial databases were an evolutionary, not disruptive technology (like Google Maps was) got me thinking though.
My own talk had also recalled the transformative moment in 2005 when Google took web mapping mainstream. Disruptive indeed, though a highly positive development for our industry. But why did the development of Oracle Spatial (which we first used at much the same time) not feel similarly revolutionary?
Was it because database technologies were better known and understood than mapping platforms among IT professionals? Did Oracle in effect slide in spatial functionality when we weren’t looking? I think this is an illusion, and that Oracle Spatial was disruptive too. The monopoly of enterprise GIS vendors was broken when it arrived on the scene, ready to power online, desktop GIS applications.
And the spatial elements still present a knowledge gap for many highly savvy database developers. My own developers and GIS analysts continue to educate each other to span the divide between the database and the spatial model. Another speaker, Andy Spears from South Gloucestershire council, agreed; his central ICT team have had to learn spatial skills from scratch to support the technology.
So is Oracle Spatial the quiet, persuasive advocate prepared to play the long game from the inside to establish itself at the heart of GIS, to Google Earth’s loud, media-hungry revolutionary, delighted to stir things up as it (literally) takes on the world? And which one do we back in the race?
Mark Thurstain-Goodwin


