Abstract

The use of Big Data, in its various forms, has gained prominence in both the private and public sectors, mainly due to its ability to provide valuable information derived from a large volume and variety of data. Among the advantages with social returns for the use of Big Data is its benefit in increasing coverage and reducing the cost of preparing official statistics, which generally rely on research involving complexes surveys and census processes to generate reliable information. The study sought an alternative to address challenges related to budget constraints and reduced response rates in census and household surveys, taking advantage of the opportunities brought about by the so-called "Information Age," using mobile phone big data and real-time location methods to observe the flow of people between municipalities to understand the dynamics of urban mobility, more specifically commuting. The levels and patterns of commuting for work or study estimated with data from the 2010 Census were compared with those estimated based on mobile phone big data in 16 municipalities in the Metropolitan Region of Rio de Janeiro. After cleaning, with the exclusion of subscribers who appear few or many times in the records, indicating that they are people passing through the region (few daily events), considering an automated dialling service (many events in the period considered), considering the existence of cell phone towers with rare events and those outside the study area, the resulting final number of subscribers considered in the study is 3,687,180. The use of such data source is not without limitations. Data loss resulting from the exclusion of inconsistent records and the loss related to the inability to infer residence and work anchors, coupled with the inability to distinguish the purpose of trips (work or study), are significant challenges. Nevertheless, with proper calibration and supplementation with other sources, mobile phone data can play a significant role in urban planning and the formulation of public policies focused on mobility. It is, therefore, a promising tool that, alongside conventional sources, can enrich the understanding of population dynamics. The results presented highlight the value of mobile phone data as an alternative source for studying commuting patterns in the Metropolitan Region of Rio de Janeiro. Despite the difference in absolute rates estimated using the two sources, mobile phone big data and census, a consistent maintenance in the pattern of movement between municipalities was observed, except for the capital, Rio de Janeiro, which showed more variation. This discrepancy suggests that local economic factors may directly influence commuting habits, revealing nuances that are not always captured by mobility analysis methods.