Dublin Bus does not release location information of its buses while in motion. However, they do periodically release information gathered from each stop a bus makes. Sometimes it can be found on Smart Dublin.
The BCD team looked at one of the days in a sample of this data added to Smart Dublin to track where, how late, and at what time of day buses were late in Dublin. Using R mapping and data table manipulation (you can see the full process on Sam Stehle’s blog post
This map uses symbology to show patterns in where the peak lateness of buses stopping are (the size of the points) and at what time of day (darker orange = later in the day). Click on the image to open an interactive version of the map
Despite the overplotting, there are patterns evident if you zoom to certain areas. The largest extent of delays – larger point radii – are on major arteries into and out of the city. There are a few reasons for this pattern. First, high-traffic roads into and out of the city will suffer from delays during morning and evening rush hours. But also, busses which serve these arteries typically will have longer routes, meaning more opportunities to become delayed. So even with high traffic in the city centre, the largest delays are at the ends of major routes.
Instead, animating the points and the extent of bus lateness throughout the day shows even more patterns in when and where buses are late.
Lateness is still represented with the radius of the point, and the time of day is still represented by color. But separating out each hour (each point is the maximum lateness of any bus stopping in the grid cell in that hour) shows that there are ebbs and flows of lateness even in the same areas served by the same busses.
Some locations have more late busses early in the day (particularly South Dublin) while others are much later in the evening rush hour (north and northwest Dublin).
You can even see when a specific bus becomes late, as a string of simliarly-sized points along a roadway stick out from the surrounding points.
You can repeat this analysis with any other date, or aggregate for a larger study period and see if the lateness patterns hold over time (though more aggregation leads to distortion of the lateness calculation – see my analysis of the SWEEP, or Temporal Aggregation Bias ).