News & Views

A big part of ITP’s work in cities in developing countries is analysing networks of buses, minibuses, shared taxis and other road-based public transport, as a first step to proposing improvements. Problems can come with loosely regulated networks of drivers competing for trade using low-capacity vehicles: in sub-Saharan Africa, for instance, motorcycle taxis and three-wheelers are common, and in some cases the largest public transport vehicles available are minibuses. As a result, some cities have few formal timetables, and little regulation of routes and services, so our work starts with collecting service data - either by ourselves or with partners. To store this data, we use the popular worldwide GTFS standard - easy to use, since it's pretty simple but also flexible.

Image Caption: Minibus departures in Monrovia, Liberia. ITP ©. 

GTFS timetables have the advantage that they can be fed into travel-time analysis tools, like Basemap’s TRACC and Conveyal Analysis. But part of our work is in trying to understand the structure of these bus networks. This can be done with GIS tools and clever visualisation, but properly understanding all the required elements of timetables, frequencies and stopping patterns can be tricky, even with the right expertise.

We'd developed tools to simplify this process a few years ago, which used clever algorithms like street routing to tell us where we thought buses were likely to travel on a road network. This is a powerful way of understanding bus networks: lots of the tools we have tried in the past look at either bus routes, or the road network, but never both at the same time. ITP’s tools estimate the roads taken by each vehicle in order to try to reconstruct the ‘whole picture’: where are buses travelling? Where are they held up by congestion? And where are they causing congestion?

Until recently, our process for this analysis generated static files that still had to be fed into GIS for interpretation. Then we started experimenting with the incredibly powerful Mapbox GL framework, which is a really flexible way of doing web-based GIS visualisation for software developers, using a technique known as vector tiles.  ‘TransitExplorer’ is the result - a bus network analysis tool, that can take GTFS data and quickly produce interactive visualisations of where buses travel from different corridors and what services are available in different neighbourhoods

The killer feature here is the interactivity enabled by the Mapbox app: where in the past, we’d have laboriously drawn static maps and worked out manually which bus routes shared which roads; now, we can now instantaneously query different parts of the road network, understanding how routes relate to each other, and where there are bottlenecks or spare road capacity. This includes important descriptive statistics, particularly frequency of vehicles and numbers of seats available. 

Image Caption: Interactively exploring and querying a GTFS timetable feed for Freetown, Sierra Leone

Developing a prototype ‘tool’ like this to be used interactively draws inspiration from hugely powerful existing transport planning tools like the Propensity to Cycle Tool, and Conveyal Analysis, both of which are brilliant for analysing cycle and bus networks in different ways, without needing loads of GIS expertise and time. The outputs can still be fed into GIS analysis tools (to check against other factors, such as land use), but it’s now possible to understand and query the overall picture without this. 

Doing more of our analysis work with interactive tools, instead of static GIS, will enable us to bring our clients, local experts, and partners into the analysis process, ensuring that their local knowledge gets fed into the questions we ask about bus networks, and helping to ensure that passenger and stakeholder requirements are thought about at every stage of the analysis process. We want to make our projects more focused on empowering local people with an understanding of their bus networks – making the benefits of public transport analysis more consistent and future changes more likely to be delivered and succeed.

For more information on how ITP's public transport analysis, including the TransitExplorer tool, can help you achieve your project aims and objectives please get in touch.