Using mobile phone data for infrastructure planning

In this paper, Dr Hadrien Salat, postdoctoral researcher and Asst Prof (Adj) Markus Schlaepfer, principal investigator of Big Data-​Informed Urban Design and Governance at the Future Cities Laboratory (FCL) programme as well as Dr Zbigniew Smoreda, present a method to estimate population densities and electricity consumption from aggregate mobile phone data in developing countries without compromising people's privacy, using Senegal as an example. The study was funded by Orange Labs.

As part of the study, the researchers used mobile phone data provided by the largest telecommunication operator in Senegal, with about 65% market share in 2013, Sonatel, which showed a low direct correlation between the average phone activity and both the population density and the night-time lights intensity. Although Sonatel is the market leader, its market coverage is not uniform over the entire country. In addition, the only population data that was available was not thorough enough spatially.

Based on these findings, the authors proposed a new method, based on the daily, weekly and yearly phone activity curves and on the network characteristics of mobile phone data, to estimate such information more accurately while allowing for the anonymity of individuals. The researches have shown that the population count of an entire census can be estimated from a substantially smaller sample of carefully selected locations.

The researchers suggest two main practical applications of using this method: a substantial reduction in traditional data collection costs and the possibility to keep track of changes in the population distribution and associated infrastructure needs

The paper external page "A method to estimate population densities and electricity consumption from mobile phone data in developing countries" was published in PLOS ONE.