Making Every Step Count:
New Model Adds Pedestrian Data to Planning Possibilities
Wes McRae | Atlanta, GA – February 25, 2026
A new model gives planners unprecedented information about walking traffic, filling a crucial gap in planning for pedestrians.
According to Spatial distribution of foot traffic in New York City and applications for urban planning, recently published in Nature Cities, the NYC Walks model presents the first large-scale pedestrian volume estimates for a major city.
"Decisions get made on data, so if you don't have data, you're essentially flying blind," said Rounaq Basu, an author of the paper and assistant professor in the School of City and Regional Planning. "We don't have good information for pedestrian traffic, so the motivation for this study was not just to make information available, but to give rich enough detail to enable planning for people walking."
One area in which rich pedestrian data changes planning is pedestrian safety. Currently, cities rely on numbers of crashes to estimate pedestrian danger. But this method doesn't account for how many people are walking there. "If we have tens of thousands of people walking on Peachtree Street, it's likely we'll end up with more crashes there," Basu said. "But when we try to identify high risk locations, we have to think of the highest ratio of crashes to people walking. We cannot do that until we know the denominator, which is the number of people who are walking, and cities do not have that information."
"The work that we're doing provides that information, and now we can get a better estimate of risk, which will inform safety planning."
Adaptable, Open-Source Model Encourages Collaboration
To create this model, Basu's team first had to build an accurate representation of the pedestrian routes throughout the city, then calibrate the model with observed traffic at different points and times of day. "We ended up creating the network by ourselves: not just sidewalks, but also crosswalks and footpaths, cut throughs and desire lines, and so on," Basu said. "All of that was modeled for the entire city of New York, which was a massive task again. The scale of effort that was needed for this paper was massive, and it's a large team. It took us over a year just to get the work done, before we even began working on the paper."
And then they gave it away.
"We shared it back with the city and made it open source. Anyone can use it, and that's always been my goal, and my collaborators share that vision as well, that our work should always be open source," said Basu.
"All the code is on GitHub, and we really hope that communities who want to plan for pedestrians in a more informed manner will be able to use these tools and create similar datasets for themselves."
"The model’s adaptability makes it a cost-effective tool for municipal planners seeking to understand foot-traffic patterns in their cities."
Want To Know More?
