Wolff and colleagues shift the study of pedestrian behaviour from isolated observations toward a reproducible comparative infrastructure. Their dataset contains 348,300 trajectories extracted from 193 hours of video across 39 European squares, collected through a common protocol and processed through object detection, tracking and georeferencing. The decisive idea is not simply scale but standardisation: comparative urban research becomes materially possible when observations from different places are rendered commensurable without erasing their spatial specificity. The study records repeated time slots and contextual parameters including temperature, wind speed, humidity, square area and city population, making the dataset both behavioural and environmental. Its methodological operation is infrastructural, converting scattered visual streams into reusable spatial evidence. The work bridges public-space research, computer vision, urban informatics and open science, while also reconnecting automated trajectory analysis to the observational traditions of Jan Gehl and William H. Whyte. It demonstrates how repeated protocols transform transient movement into a durable comparative archive.
URBANISM MEETS ART