Redefining Urban Landscapes: A Methodological Approach to Transforming Underused Parking Spaces with Dynamic Urban Functions

Abstract

This study presents an approach to identifying underutilized urban spaces, focusing on parking areas, and explores potential reutilization strategies in Greater Boston. Under the milieu of the information age, global urbanization, and technological development, the prosperity of urban data serves as the new method to approach urban proposals. The city, as a multifaceted artifact, is examined through the lens of advanced data-driven techniques, particularly deep learning. With the computer vision model, the underused surface parking lots will be automatically detected according to historical satellite imageries, highlighting a misalignment between the current infrastructure and the actual urban needs. This study then leverages miscellaneous urban factors to analyze the parking patterns. Associated with the multimodal system, there are possibilities underlying the usage of redundant surface parking. Considering the high rents and housing situation, these spaces could be transformed into housing units or even mixed-used districts, to alleviate the housing crisis in Greater Boston.