Automobile accidents happen every minute of every day, and they are one of the leading causes of injury and death in the US. In 2016, an average of 102 people died every day as a result of a car crash. Throughout the years, safety measures like seat belts, airbags, and AI have helped dramatically in reducing fatalities, but not nearly enough.
We wanted to gain an understanding of these accidents and use a data approach to help solve some problems. We were really inspired by NYC Open Data, so we wanted to explore the dataset of all car collisions since 2013 and potentially learn some cool things.
What it Does
It lets users view and interact with a visualization of NYC car collisions.
How I Built it
First, we worked to access the database of motor vehicle collisions, which is accurately accounted for by the NYPD and can be found on the City of New York page. We converted the data from its raw form into a condensed sheet for analysis and into a json file to work with D3 and numpy. Finally, we overlayed files of the locations of accidents on top of the map of NYC.
What I Learned & Challenges
It was difficult to parse and find/modify a geojson file for the visualization. We learned a lot about how to use D3 and geojson.
What's next for Visualizing NY Car Crashes
We anticipate broadening our visualization to include other factors like car types and the reason for the crash. Also, adding a simulation of accidents throughout time would be interesting and expanding to other cities and areas.