Improving Community Safety and Economic Well-being by Remediating Buildings with Roof Damage in Baltimore

Fellows: Justin Clark, Jonas Coelho de Barros, Chae Won Lee
Data Science Mentor(s): Kit Rodolfa
Project Manager: Abiola Oyebanjo
Project Partner: City of Baltimore

The city of Baltimore, MD, has seen a rapid decline of population over the past several decades, leaving thousands of vacant buildings as a consequence. As the condition of these buildings deteriorate, they can have an adverse impact on the structural integrity of adjacent residences (especially among row homes that comprise the majority of housing units in the city) as well as on the economic and community wellbeing of the residents. Without any intervention, these buildings pose a risk of collapse with tragic consequences: three firefighters recently lost their lives when a vacant building they responded to collapsed. Baltimore’s Department of Housing & Community Development has partnered with the DSSG fellowship to help them identify hazardous buildings with roof damage to prioritize for preventative interventions. To do so, our team is combining aerial imagery with data on historical inspections and 3-1-1 calls to develop machine learning models to detect buildings with extensive roof damage for the city to inspect for possible emergency demolition, stabilization, or other remediations that can benefit local residents and communities.