Archive for March, 2013

Reflections on OpenStreetMap

March 9th, 2013

The second Canadian OpenStreetMap (OSM) developer event held at Ryerson’s Geography department started today with a series of presentations and workshops introducing students and members of the broader community to OSM. Toronto OSM guru Richard Weait gave another one of his engaging OSM-or-nothing speeches, telling tales of trap streets and mappy hours. He also got attendants to edit the OSM data and submit a few new features based on their local knowledge of their neighbourhoods or the university campus. Geographic Analysis student, GIS consultant, and spatialanalysis.ca blogger Michael Markieta guided us through the querying of the OSM “planet file” from a PostGIS/PostgreSQL database and its mapping in the open-source Quantum GIS package (see photo).

michael-teaching-osm-queries_08march2013

As most of you will know, OSM is a global volunteer project to create a free geographic base dataset. OSM data have been shown to be more detailed and accurate than commercial data, at least in some areas of the world. There was some interesting discussion this afternoon about potential liability issues due to inconsistencies in OSM data used in professional applications. The concern that OSM contributors could be held liable for erroneous contributions was countered by noting that commercial data vendors provide their data “as is” in just the same way, and that their data are out-of-date most of the time. That certainly seems to be true for my car navigation system! Still, the possibility of downloading OSM data for a professional map at a moment where a misuser has modified or deleted information that has not been detected and reverted by the community makes me uneasy. Also, the thought that detail in OSM, e.g. in rural areas, may depend on whether or not there is an avid mapper living in the area, is unsatisfactory.

Further, the challenges resulting from free tagging of new features were brought up at today’s event. There are support sites such as taginfo.osm.org and the map features list on the OSM wiki, but I cannot help but think that the OSM community is repeating mistakes that were addressed (at least to some degree) by research, development, and best-practice in GIS over the last couple of decades.

Whatever your position with regards to these issues, OSM is playing an increasingly important role in government and business. Our students need to know about it, and I think today’s workshops went a long way to achieve this awareness. Thank you to Mike Morrish and the Student Association of Geographic Analysis (SAGA) for their tremendous support in organizing this educational event and for sponsoring food and drinks today.

From a research perspective, OSM is a fabulous subject too. My interest in it was discussed in a section of an earlier post about volunteered geographic information (VGI) systems. The OSM developer weekend is focusing precisely on hardware, software, and provider/user issues that are not well explained by the VGI label, but captured within our concept of VGI systems to be presented at the 2013 AAG conference.