Poindexter Village was one the first public housing projects in the United States; Franklin Delano Roosevelt attended its dedication in 1940. The village was named in honor of community leader, Rev. James Preston Poindexter and became a close knit community for many in Columbus. Although most of the buildings have since been demolished, the two remaining buildings will be part of the new Poindexter Village Museum and Cultural Center. 

Mt. Vernon Ave was the commercial heart of the Black community in Columbus in the mid-20th century; the construction of the I-71 highway severed this corridor from Columbus downtown. We are initially focusing on a three-block stretch (from present-day Monroe St. to 20^th St.) that had the highest concentration of commercial activity in 1951. The buildings in this area were demolished and some streets removed as part of an urban renewal project in the early 1970s. 

Hanford Village was originally established in the early 1900s and became home to many Black residents during the Great Migration. From 1943-1946, Hanford Village’s George Washington Carver Addition was built. This project developed 146 homes specifically for Black veterans of World War II, including Tuskegee Airman. Unfortunately, less than 20 years later, 87 homes—including 67 from the Carver Addition—were relocated or demolished for the construction of I-70. Today, the neighborhood is still divided by the highway.

Sanborn Maps were hand-constructed for US cities across the country for fire insurance purposes and date back to the 19^th century. The maps contain a wealth of building level data including building footprints, uses, construction material, number of stories and more. An image of a Sanborn map and accompanying key can be found below. CURA has developed machine learning techniques to automatically extract building footprints and other key information from Sanborn maps. 

By creating a GIS database through this machine learning process (example below), we’ve opened the door to building 3D visualizations. CURA researchers use the building footprints and number of stories to create basic 3D models in Esri City Engine.

For our focus neighborhoods, CURA researchers have used historic photographs, street imagery, archival data, and memories from community elders to build realistic geometries and textures (building exteriors) for key buildings. Ongoing community engagement will lead to further improvements in the models' accuracy and immersion.

CURA is working toward increasing the scalability of our methods. We're hoping to extract additional data from Sanborn Maps, such as windows per floor and building addresses, to build more detailed initial models quickly. Once we achieve greater efficiency in our processes, there are broad applications for this work:

• With more plausible initial models, there is a stronger foundation for co-design with communities
• With neighborhood or city-wide models, we could create high-resolution urban morphology databases to conducts spatial analyses of the built environment over time
• With building addresses extracted, we could automatically incorporate data from city directory data or other online databases