Geo-demographics from CDR Data: Mobility
Understanding mobility at a population level provides a key role in urban planning. Traditionally undertaken by transport surveys, a key outcome is city wide Origin-Destination (OD) Matrices. While providing known utility, these traditional approaches are both time and resource intensive. Within many developing countries these costs, coupled with the rapid nature of urban grown, mean these surveys are simply not undertaken with urban planning suffering as a result.
Addressing this issue the N/LAB has developed and demonstrated a methodology for repurposing billing record based CDR data (while companies may record hand-offs events enabling full movement tracking this cannot be assumed) to automatically extract equivalent absolute valued origin-destination matrices. Working with the World Bank this approach has been implemented within Dar es Salaam in Tanzania, directly providing impact to urban planners.
As part of the work two technical reports have been developed, with the first openly available and the second available on application. The first details the technical development methodology of Origin-Destination (OD) Matrices from CDR data, including:
- the methods of their creation and validation
- a critical assessment of the state-of-the-art in OD Matrix creation
- a discussion of the theoretical underpinnings of that process and the challenges
- an agenda for future research and development that may serve as foundation for overcoming challenges in using CDR data for mobility analysis, and to fully realise the extensive opportunities it opens up.
The second details the insights and findings within the specific city, providing fine-grain maps and summary statistics that illustrate the population, mobility, accessibility of public transport networks and Origin-Destination densities.