The aim of this study is to understand how mobility behavior patterns, extracted from spatial and temporal urban data such as Replica Data, can be applied in a simulated environment. The simulation, conducted using the CityScope platform, allows for testing various urban interventions and visualizing their impacts on individual mode choices, along with the resulting second-order effects. Focusing on Dallas as a case study, our research aims to predict how transportation mode choices could become more sustainable with changes in inner-city land uses.
Employing a binary mobility choice classifier, the study models these choices based on characteristics extracted from areas exhibiting desired land use attributes and rider behaviors, including trip origin-destination coordinates and demographic information. The research underscores the significant influence of specific urban interventions in fostering walkable communities, such as housing densification, workplace creation, and amenities and services within walking distance.
These interventions demonstrate a paradigm shift, with a significant 71.2% of the population favoring 'green modes' of transport, providing actionable insights for future urban planning strategies in Dallas and potentially other cities. Keywords: predictive modeling; urban planning; transportation.
Acknowledgment: This research was conducted in collaboration with NTT DATA and supported by data provided by Replica and SafeGraph.