Machine Learning from Las Vegas

Urban Form Performance Analysis using Temporal Telecom Data

Urban environments are inherently temporal by definition: they contain forms [buildings, public spaces, streets and landscape] that are staged to allow activities [people movements, traffic or commerce]. Nevertheless, evidence based analysis of urban forms has always set a challenge for the design and planning community. The lack of tools and data that can shed a light on how human flow is affected by their designs left many design decisions unexplained or unproven.

With the constant emergence of new spatial analysis methods and availability of massive temporal data sets [‘big data’], researchers can now better expose behavioral patterns of crowds within dense urban settings. This paper aims to explore a method to analyze patterns of usage for large, random groups within urban settings using Radio Network Controller [RNC] records. By matching behavioral patterns such as clustering, stay and persistence with spatial parameters such as urban-form and amenities, this work aims to find evidence-based correlation between the design and utilization of urban spaces.

As a case study, this work focuses on the city of Andorra La Vella, Andorra. This rather small city becomes a regional tourism hub for several months over the year. The large and diverse population of tourists poses a challenge to understand patterns of usage of urban spaces as their patterns are naturally less repetitive and more random. By obtaining, distilling and analyzing temporal RNC data of 2016-2017, we’re able to observe the activities of both tourists and locals within the city. Using clustering methods we extracted areas that are more prone to attract these users to ‘stay’. 

Powered by