Project Description
Examples of work, during my tenure as Co-Director of the project: "Developing socio-spatial clustering algorithms to unveil urban dynamics from near real-time non-conventional data sources for urban planning, management, and monitoring of cities". Tasks included the development of the proposal and supervision of the project for a year. Credits for collaborators: Horacio Samaniego, Boris Sotomayor, Mauricio Cisternas (UACH), Pablo Valenzuela, Fernando del Campo, Sebastián Pereira, Amanda Ibsen (INRIA).
Proposal and Methods
The initial proposal consisted in the exploration of several possibilities of Urban Boundaries as defined by socio-spatial clustering algorithms (based on the spectral analysis of networks), applied to mobile phone communications data in Chile. By summarizing in near-to-real time properties of cities such as Population, Infrastructure and Production, UACH researches were expected to a) find a proper Urban Delineation for chilean cities, and b) detect critical deficits in urban infrastructure (according to the theory of Urban Scaling), which could be useful for urban and transportation planners.
Proposal and Methods
Vision for Wall-Display visualizations.
Proposal and Methods
Methodology for constructing urban clusters.
Proposal and Methods
Mehtodology for exploring urban boundaries and summarizing citiy properties.
Big-Data solution
Architecture of Big-Data infrastructure, used for applying socio-spatial clustering algorithms to mobile communications data, and visualizing results in near to-real time.
Wall-Display
Testing Wall-Display with results from clustering technique applied to the Santiago-Valparaíso region.
Wall-Display
Testing the application of clustering techniques for the Santiago-Valparaíso region, on data from government’s Origin-Destination surveys.
Web-Interface
Demo of Web-Interface (Developed by INRIA engineers), which provided near-to-real time aggregates for Population, according to several agglomerations of geographic entities (4 distinct units), from several clusters as defined by a slider based on a simple clustering technique (cell phone activity).
Web-Interface
Left: Example of population aggregates for urban clusters at specified threshold level, Right: Chile’s Cell Phone Towers (30K+) Map.
