The collective mission of mapping the world is never complete without street addresses. Crisis response as well as longer term rehabilitation requires up-to-date maps including street names and house numbers to geocode physical locations. We describe how to enhance crowdsourced OSM road geometries by generating automatic addresses so that aid agencies and locals can better coordinate location-specific services. For example after a crisis, OSM mapathon community remotely creates road vectors. While the actual naming may take longer, an algorithmically generated scheme provides a common framework to start immediate coordination.

Our system processes OSM road vectors and outputs temporary labels for streets and houses in any unmapped area with missing labels. Our generative algorithm creates hierarchical and linear street addresses (Robocodes), by labeling regions, roads, and blocks; based on addressing schemes around the world and coherent with human cognitive system. For the algorithmic details, we invite the audience to check our award winning CVPR EarthVision paper (at robocode.info). In this talk, we will share and demonstrate improved maps on ID-tool, compare them to current open and industrial solutions, and walk through our opensource code to generate Robocodes for any given OSM bounding box.