Applying Big Data Tools and Techniques to OpenStreetMap with OSMesa

Sunday, 9:30 am – 20 min
Seth Fitzsimmons
Pacific Atlas

OSM contains an enormous amount of data, even more when one considers the historical footprint produced by 14 years of editing. As a community, we’ve generally addressed the problem of surfacing it in maps, but there’s much more that can be gleaned.

How have particular areas evolved over time (both as mapped and physically developed)? How complete is the map? How have Missing Maps, HOT, and other guided mapping programs influenced it?

There are some roundabout ways to get at these answers, but it’s never been particularly clear, comprehensive, or quick. OSMesa is a set of open-source tools that enable us to easily explore OSM to tackle these questions and help us better understand the map and the world around us and to empower the OSM community to tackle ungainly projects in more informed ways.

In this talk, we’ll discuss OSMesa’s underlying philosophy and technology (Apache Spark, GeoTrellis) and walk through how it’s used, including to support the Missing Maps leaderboards and for assessing completeness in African nations. We’ll share the vision for the project and elaborate on how this approach to working with OSM data can dramatically expand our ability to ask and answer questions.