Physical trace on map from GPS track data

gps trackWith more and more people carrying personal tracking devices, a wealth of personal information is being collected that can be visualized or physicalized in different ways. The geospatial data that all of our phones – and more and more watches – collect is particularly amenable to rendering into physicalizations.

Below is a model of the path I took a few months ago as I ran a few errands around my home on Aquidneck Island. I made two stops and intentionally took a route that formed a loop with no overlapping segments.

I collected the data using trails, a GPS tracker for the iPhone. After exporting the data, I used it to create a physical graph of the trip with x and y coordinates representing the path I took and the height of each segment showing my speed at that time.

I made two. One with the entire trip printed in one color

 

single color track

 

And another with the three legs of the trip shown in different colors

multiover

 

Encoding speed as the height of the path highlights some of the features of the trip. I drove much more slowly on the small sides road in my neighborhood compared to the larger roads on which I spend most of the trip:

neighborhood speed

 

The many stop lights I hit are also apparent as deep, short dips in the line.

Another feature of the trip that stands out are three segments along the larger secondary roads where I slowed down, but didn’t come to a stop like you might expect if I had hit heavy traffic. I took this trip on a weekday while school was in session and those slow segments indicate the three school zones I drove through along the way.

As usual, the photos don’t capture the experience of interacting with the physical object. Completing this project convinces me that – as with other physicalizations I’ve made – ones based on personal tracking data have the potential the deepen understanding and increase engagement with information.

It will be fun to continue to explore physicalizations like this. One possible extension would be to use exercise data from a runner or cyclist that includes more than just speed data. Most phones can capture elevation as well as speed and if a heart rate monitor was used that data could be included as well. With a richer dataset the model could have multiple lines, each one representing different aspects of the workout.

 

onesegment