I had the honor of giving the E. Willard and Ruby S. Miller Endowed Lecture in the Department of Geography at Pennsylvania State University on March 15, 2024. My talk was titled “Mapping Columbus’ Ghost Neighborhoods: Using AI and GIS to Create 3D Models of Neighborhoods Damaged by Urban Highways and Urban Renewal in the 20th Century.”
They even baked a cake!
New paper: Stiles, J. and Miller, H.J. (2024) “The built environment and the determination of fault in urban pedestrian crashes: Towards systems-oriented crash investigation,” Journal of Transport and Land Use, 17, 97-113.
Abstract: This study identifies built environmental factors that influence the determination of fault in urban pedestrian crashes in the United States, with implications for both safety and equity. Using data from Columbus, Ohio, we apply regression modeling, spatial analysis, and case studies, and find pedestrians are more likely to be found at fault on fast, high-volume arterial roads with bus stops. We also observe that better provision of crossings leads to more marked intersection crashes, which are less likely to be blamed on pedestrians. In addition, large differences in both the provision of crossings and fault exist between neighborhoods. We interpret findings through the lenses of the systems-oriented safety approaches Safe Systems and Vision Zero. The conclusion argues that the designation of individual responsibility for crashes preempts collective responsibility, preventing wider adoption of design interventions as well as systemic changes to the processes that determine the built environment of US roadways.
Media
New paper: Lin Y, Li J, Porr A, Logan G, Xiao N, Miller HJ (2023) “Creating building-level, three-dimensional digital models of historic urban neighborhoods from Sanborn Fire Insurance maps using machine learning.” PLoS ONE 18(6): e0286340. https://doi.org/10.1371/journal.pone.0286340.
Abstract. Sanborn Fire Insurance maps contain a wealth of building-level information about U.S. cities dating back to the late 19th century. They are a valuable resource for studying changes in urban environments, such as the legacy of urban highway construction and urban renewal in the 20th century. However, it is a challenge to automatically extract the building-level information effectively and efficiently from Sanborn maps because of the large number of map entities and the lack of appropriate computational methods to detect these entities. This paper contributes to a scalable workflow that utilizes machine learning to identify building footprints and associated properties on Sanborn maps. This information can be effectively applied to create 3D visualization of historic urban neighborhoods and inform urban changes. We demonstrate our methods using Sanborn maps for two neighborhoods in Columbus, Ohio, USA that were bisected by highway construction in the 1960s. Quantitative and visual analysis of the results suggest high accuracy of the extracted building-level information, with an F-1 score of 0.9 for building footprints and construction materials, and over 0.7 for building utilizations and numbers of stories. We also illustrate how to visualize pre-highway neighborhoods.
Media
