Mapping Mobility

One requirement of the ENGL 229 is a “Wildcard” blog post, which is similar to the other course assignments with the only difference being the subject choice lies at my discretion.

As such, for my wildcard project, I decided to assuage my curiosity, peaked at a very interesting column header in the metadata collected on the Fulcrum mobile app, and that is the “GPS_SPEED” column. I already know that GPS modules in cellular phones can calculate speed, which is how Google generates its traffic maps, however I didn’t know that speed and direction were recorded into the EXIF data of an image alongside the static coordinates.

This gave me the idea of generating a mobility map to try and establish a certain cultivated standard on roads, where the effect of regionality I expected would be lower since they are travelled by a larger and more diverse chunk of the population.

However, a cursory look at the data at hand reveals that it is far from clean or standardized. There are more than 600 entries which are blanks, and 600 more which are tagged as -1. Looking online, I couldn’t find any indication on what this could mean, however the general consensus was that the EXIF data tagging mechanism varied wildly from hardware to hardware, with many irregularities often seeping into the GPS section of the data (as witnessed during my own datacollection, with automatic geo-referencing often missing the mark by kilometers)

Nevertheless, I carried on and created the following map:

For this map, I considered the "-1" entreis as being static alongside the "0" ones, and did not include any of the blank points. Another thing I attempted was the generation of two superimposed animated heatmaps, to see of there were any different patterns between data collection on foot and in a car, however cartoDB didnot seem to support multiple animated heatmaps as layers, and a map with two static heatmaps was too noisy and not useful.

There is a lot of unexpected overlap on main road networks, which could be either attributed to taking pictures while in traffic or to the aforemetionned artefacts and general iffiness of the EXIF geo-tagging mechanism. There are however some clear emerging patterns: The bulk of the collection around the center of beirut seems to be static, while there are some streaking and usually linear moving data points on major roads outside of the city.