New Cesium showcase: Divvy.Vision


Check out the latest Cesium showcase app:

Divvy.Vision is an open-source viewer of 2014 origin and destination trips of Chicago’s Divvy bike share program, and an entry for the Divvy Data Challenge 2015. It was developed by Matthew Shaxted using several technologies including Cesium, CartoDB, D3, and Turf.


The arcs do represent bike rides and not Joshua playing a Thermonuclear War simulation, correct? :smiley:

Kidding aside, this is a really cool program. The wire-frame buildings is a nice touch. The white wireframes with a black background makes for good contrast. Do the bikes have GPS tracking for exact routing? If they do perhaps showing that data might be too much information, even though the rider identity is anonymous. So a tall but skinny curve indicates that the origin and destination are not far from eachother, but the rider took a long round-a-bout path in between. This program is an awesome way to quickly visualize large amounts of data over time.

Thanks - Cesium can handle a lot of building geometry quite nicely actually. Great platform for this type of city visualization work. That video clip reminded me of this visualization:

Divvy bikes currently do not have GPS tracking for exact routing. When a customer checks out and checks in a bike to a station, they record a flag, and they make these flags available. Each data point has station id and timestamp of checkout, trip duration, type and age of rider if registered among a few other things.

Would be cool if they were all GPS enabled though... could probably infer a most probable route by looking at trip duration and comparing to known routes.


Matt’s app just made it to the finals for the Divvy data challenge. Please consider voting for it:


Looks like the only one with 3D visualization, also the only one with streets marked (with buildings to boot), unless I’m missing an option in the other ones. The moving dots seem to simply move along an L shape, while it appears the one with lines move along possible routes along the streets, though it doesn’t show the streets on the map that they are following. I think I prefer E.T.'s routing!