OpenStreetMap is a giant datastore, with an extremely flexible data model. Its API accepts all additions. But, in fact, not everything belongs or fits in OpenStreetMap.

Some thematic data require more advanced modeling than OpenStreetMap’s simple tagging scheme supports. To represent a building in full 3D requires a data model that supports solids, and to represent a public transit network or traffic patterns requires a data model that handles space and time.

Other types of data come from authoritative sources and may require cleaning, combining, and perhaps even legal review before they’re ready to be added to OpenStreetMap. For example, street addresses and trailheads.

Just as not every kind of data can fit into OpenStreetMap, not every kind of data can easily be extracted from OpenStreetMap. Storing data outside of OpenStreetMap proper may also make that data more relevant to other users’ needs. For example, supporting data consumption in mobile apps, or supporting data collection with topic specific editor apps.

We’ll discuss peripheral data to OSM, both in terms of technical implementation and in terms of community impacts. If done well, data projects that are connected but not subsumed by OSM can advance open geo data. Let’s figure out together how that’s best done.

Panelists:

• Ian Dees, OpenAddresses
• Jan Erik Solem, Mapillary
• Jereme Monteau, OpenTrails
• Drew Dara Abrams, Transitland
• Peter Richardson, PolygonCity