Access to Prosperity: Quantifying Infrastructure Impact With OSM
2019-09-23 , Großer Hörsaal

In many regions of the world, a population’s access to essential services is unduly constrained by a lack of proper infrastructure. By performing accessibility analysis using OSM data, we can understand how route infrastructure impacts access to essential services and use that information to inform an intervention.

This talk explores accessibility analysis performed to understand the impact of footbridge construction in eSwatini and introduces a python framework enabling users to perform similar analysis.


Bridges over rivers are fundamental pieces of human infrastructure which enable safe crossings for the populations who use them. However, many rural regions of the world lack these critical bridges, and as a result, access to essential services is restricted for millions of people.

In pursuit of tackling poverty caused by rural isolation, we have begun experimenting using remote analysis techniques to quantify bridge need and impact. To these ends, accessibility analyses using OSM and other data have proven to be extremely valuable.

By comparing a population’s baseline accessibility to its access in scenarios such as a flood or post bridge construction, one can better understand the transportation dynamics of a region and even estimate the number of people that would be impacted with a bridge construction, for example.

During the course of running various accessibility models, a python framework tebetebe was built around OSRM [Open Source Routing Machine] which simplifies the process of running different scenarios. This framework is built with footbridges in mind but is generalized so it may be applied to other applications.

This talk explores the accessibility analysis process, its caveats and results achieved in the Kingdom of eSwatini. Finally, a walkthrough of the python framework tebetebe is given so that other users may perform similar analyses.


Talk keywords:

access,ngo,africa,routing

See also: slides (30.6 MB)