Urban Data Science & Equitable Cities

There is a growing conversation in academic circles, media, and decision-making spaces about the need to achieve gender equity in many aspects of society, including transportation. In urban transportation, public transport safety concerns disproportionately affect women, while transportation systems for all modes (transit, bike, pedestrian) are still not implementing policies and solutions that sufficiently adapt to women’s needs. One challenge to designing effective policies is a general lack of disaggregated, reliable, and up-to-date information about women’s travel patterns. This talk will focus on new avenues to explore differences in travel patterns between men and women and how these results could translate into urban policy, using Canadian cities as case studies.

App-based ride-hailing platforms have transformed urban mobility, but their expansion has raised growing concerns about equity across neighborhoods. This talk brings together two projects to examine this topic with empirical evidence and methodological approaches. The first project examines how ride-hailing platforms and traditional street-hail taxis have served different communities over time. Using six years of New York City trip records, I analyze spatial-temporal differences in pricing and coverage between Transportation Network Companies (TNCs) and taxis. The results show that these differences are dynamic and became more pronounced during major disruptions such as the COVID-19 pandemic. Areas with higher shares of carless households experienced rapidly widening fare gaps, while minority-concentrated neighborhoods initially benefited from lower TNC fares but gradually lost relative coverage advantages during and after the pandemic. The second project shifts from measuring disparities to addressing them with predictive demand modeling. We show that common spatiotemporal prediction models prioritize overall accuracy while overlooking spatial and demographic imbalances in prediction errors. To address this, we introduce a Residual-Aware Attention block and an equality-enhancing loss function that explicitly account for spatial disparity during training. Applied to travel demand prediction in Chicago, this approach substantially reduces spatially clustered errors and improves fairness metrics with only modest losses in accuracy, supporting more equitable planning and policy decisions.

Many cities today are redesigning their streetscapes to redress the historical privilege afforded to the automobile in planning and policy. Much streetscape redesign is around transport infrastructure space, which largely prioritizes car travel and marginalizes other travel modes. Attempts by planners and policy makers to this end often are met with public opposition by advocates of the car, protesting about losing space on the street. This is empirically investigated with the case of Montréal by determining the allocation of street space to transport infrastructures, deriving measures of infrastructure space per traveller, and devising an Equal Infrastructure Allocation score to measure the imbalance between infrastructure provision per travel mode. Per borough, the distribution of transport infrastructure is examined, alongside correlations with demographic, socio-economic, land use, and crash rate variables. Potential scenarios of significant micromobility infrastructure improvement are modelled to test how infrastructure space apportionment per mode changes. This investigation discovers that even large improvements to micromobility infrastructure have a minor effect on space allocated to automobiles. Equal Infrastructure Allocation score and associated indicators are presented as useful tools for planners and policy makers implementing micromobility infrastructure projects, to better communicate with the public and address potential opposition.

Previous studies on urban heat mitigation, critical for urban planning and public health, have generally focused on a handful of cities, ignored logistical constraints, or insufficiently resolved urban-scale processes. Here, we fuse satellite-derived estimates of urban heat and multiple physical properties to develop a non-parametric machine learning approach to capture non-linearities in thermal anomalies (ΔAT) across 493 U.S. cities. This enables computationally-efficient data-driven assessments of urban heat mitigation strategies, including strategies targeting low-income communities since ~90% of these cities show income-based temperature disparities. All strategies lower daytime ΔAT, with targeted afforestation with (without) albedo management reducing daytime ΔAT for low income groups from 0.56±0.94℃ to 0.22±0.92℃ (0.24±0.93℃) and income-based ΔAT gap from -0.50±0.94℃ to -0.15±91℃ (-0.17±93℃). Our results demonstrate the importance of targeted heat mitigation in low income communities, where residents have less options to adapt to extreme heat.

Presenting fun maps, forming mapping groups, setting up a gameplan for the 30 Day Map Challenge (running during the month of November), and walking members through our GitHub repository for the challenge.