Data Lab: Why is my bus late?
Overview
Need better understanding of data sets, what they can do.
This includes an understanding of the drawbacks / limitations as well as the benefits and insights that are possible.
How can data be harnessed and interpreted to give better insights?
This document focuses on the bus delay insights that can be derived from DfT’s bus open data API.
Why?
Measuring bus delay very important because it highlights where interventions are needed the most.
Should we focus on passenger loading - perhaps improved ticketing, two doors? Or better revenue collection to minimise the interaction time as passengers board buses.
Or different vehicle designs: single deck vehicles - double doors.
Or should our interventions focus on pinchpoints on the route. Are there certain locations where buses habitually get held up - traffic light timing - bus lanes. etc
target bus lanes where they are needed. don’t want to reallocate road space if this intervention isn’t necessary.
better insights on time location, routes, vehicles, variation
We can’t answer these questions and target interventions in the most cost effective way, without a comprehensive understanding of bus delay.
this map shows the locations of three random ‘pings’. @cr-pings-random
each pings represents a bus telling us it’s location. @cr-pings-random
As well as the time and location of the ping other attributes including: - bus service number,
- destination and
- direction
In just 1 minute, there are almost one and a half thousand ‘pings’ from buses in the West of England area.
Each bus on the network sends a ping roughly every 20 seconds,during the day, the number of pings continually increases.
… in just 1 hour buses will have produced about 80 thousand pings.
This gives us a good overview of where buses are running. We can clearly see concentrations of pings on routes emanating from the city centres. Conversely the concentration of pings is much sparser in more rural areas such as the Chew Valley.
Challenge to collect, store and format the large number of pings.
Let’s focus on one route.
These are the pings for the 1 bus service. The route goes from Cribbs Causeway to Broomhill (Brislington).
Even in this short period (30 minutes) there are lots of pings, so it’s hard to discern any patterns in the data.
Let’s see the pings for just one unique bus journey.
Animating the pings for this trip gives us an idea of the speed and where the bus is being held up during its journey.
We can also represent the bus journeys on a chart.
The morning bus journeys have been highlighted.
This chart shows total journey times for several bus journeys. Variation. Peak times.
So, what are bus pings?
