
Open Street Map is a fascinating project. The idea is to build a 'open source' map of the world's streets based on UGC (User Generated Content), and make this data available free of charge, as an alternative to the fee-based maps from Navteq and Tele Atlas. Interesting, yes. How is it doing?
Well, at first look, things seem promising. Bringing up the map of the USA on the OSM site shows all sorts of roads, and no major holes. But, things are not that simple.
The VAST MAJORITY of the data in the OSM road network in the USA is from the (also free) TIGER/line database, provided by the US Government Census. And although this data is OK from a road vector perspective, it is not fully complete vector-wise, and more critically, lacks much of the additional attribution (max speeds, directionality, and turn restrictions) that car navigation solutions require. Generating that attribution would be slow and costly. The UGC aspect fails here, as who wants to spend a weekend doing that?
Where things are MUCH more interesting are on the trails/walking side. Here, the open, UGC-friendly side of OSM has a huge advantage over Navteq and Teleatlas. On a daily basis, people are walking trails with GPS units, logging their paths, and uploading it to OSM. Quick, fun, and it scales. There, it becomes part of the known OSM world. Navteq's Discover Cities product, which also includes trails, are gathered by Navteq employees and entered into the Navteq database the same way as roads are. It's a slow, costly process, and does not scale.
I'm not sure OSM has a play in Autonav, where all of the complex road segment attribution is missing. But, as pedestrian navigation becomes more important in the LBS/GPS world, I see a bright future for Open Street Maps.

1 comments:
Hi Clay,
Thanks for this info about TIGER and Open Street Map.
Do you know what TIGER year has been imported into OSM and whether there are plans to update OSM as newer TIGER data become available?
John Williams
Post a Comment