Thursday, April 29, 2010

SafeRoadMaps - Visualizing Car Accident Data

The visualization that I want to present this time is from SaferRoadMaps.org, a research project that was developed at the University of Minnesota to help visualize data that was collected about traffic accidents in a geographic sense. This system is powered by Google Maps and augments that functionality with additional filters and data to provide users with different kinds of insight into traffic accidents. Probably what interests me most about this visualization is the level of interactivity you have with it to find data. More on this after the photos:


Pictured above is a "crash density" map. The brighter the dots, the higher the collision rate. I depict Mainland U.S. here, but there are maps available for Alaska and Hawaii as well (which are also zoom-able to street level made possible by Google Maps).



Above is the same type of map zoomed-in over UCI. You should be able to click the picture to get a better look. At this level of detail, this map is similar to the Dr. John Snow's cholera map covered in lecture.



And the best for last, an example picture of what kind of data that can be augmented. You can click on accidents using the SaferRoadMap's "View Street Maps" tool to see data about any fatalities, ages of the people involved, if they have deceased, if speeding was involved, if drinking was involved, and the type of road it occurred on. This is one of the beauties of computerized visualization, as you can pack in this amount of data and still make it accessible on limited sized screens. Technology has also made it much more easier, if even possible, to aggregate and combine all these levels of data into one visualization.

To explain the filters that I have mentioned a little earlier (but not in the picture), you are able to directly type in addresses to auto-zoom to a relevant area on the map. You are also able to filter for driving laws (aggressive law, seat belt enforcement, etc.), road types, date types (summer or non-summer), speed types, drinking types, and restraint types. You are also allowed to filter results for people involved in collisions: Gender, Age, and if the person was a bicyclist, driver, passenger or pedestrian. Filters are accessible from the "View State Maps" link.

No comments:

Post a Comment