Sonic Urban Transformations: A computation model to study and represent temporal changes in the walking experience.

Thesis Advisor: Terry Knight
Readers: Arvind Satyanarayan (EECS), Andres Sevtsuk (DUSP), George Stiny

Cities are dynamically changing, complex environments, especially during unpredictable events like the global pandemic where parking lots and sidewalks evolve to become restaurants at certain times of the day. Yet, the current urban models used by urban planners and designers include only static representations of the city, that rely on visual information such as maps and images. These static representations of the city are incapable of capturing, representing, and accounting for the changing condition of cities and peoples’ lives. Thus, urban design and planning decisions remain insensitive to the social and spatial conditions that are in constant flux.

Urban spaces are ephemeral, temporal, and ambiguous in their nature and that they are best perceived in motion and through time. The thesis, I propose, forms a computational model to understanding and representing the temporal changes in the walking experience through sound. Sound offers a more dynamic representation of everyday life in the city as it can convey information about the changes in the practices, actions, and events that take place in the space. These temporal changes in the practiced space constitute not only spatial transformations but also sonic transformations that shape the walking experience. The development of a novel system of representation to enable the effective comparison between different walking experiences is a key component in this work.

A specific walking route around Harvard Square was used as a case study to capture how the different phases of the pandemic changed the walking experience at street-level and the spatial conditions over time. By collecting visual, audio data and geolocation data, captured over a three-month period in the afternoons and evenings, the comparative model was built. Urban planners can use this model to understand how the planning decisions affect the walking experience and inform their decisions by the temporal multimodal representations of city life.