Crime Watch

2013 (completed)

Crime Watch was created in ~20 hours as part of a homework assignment that used to be a requirement to join Ringtail Design. The homework described a San Francisco Police Chief needing to identify patterns, trends, and outliers from a 10,000 entity KML file so she/he can plan police coverage for the upcoming months. I implemented a responsive web application and created some reveal.js slides to present to the hiring team.

The Solution

An application was created to answer 3 major questions

Histogram

A histogram identifies the top types of crimes:

Geospatial Clustering

Numbered clustering identifies the areas with the most crime, which is highly concentrated in the courthouse area:

Timeline

A timeline identifies trends in crimes, most notably a peak in the middle of August and a slow decline leading into October.

Mobile and tablet support

Since the app was responsive, it was fairly easy to port it as hybrid iOS and Android apps. For the demo, I brought along my iPhone, iPad, and an Android phone.