BUILDING A STREET DETERIORATION MODEL TO PREDICT THE FUTURE QUALITY RATING OF STREETS

Client
New York City Department of Transportation
Faculty
Davenport, Geoff
Team
Michelle Mak Ng, Zhenli Yan, Yulin Chen , Tianhao Wu, Shumao Huang

The New York City Department of Transportation (DOT) is responsible for maintaining 6,000 miles of streets and highways within city limits by prioritizing the funding of street reconstruction work based on a 10-point quality rating system. DOT leverages a pavement deterioration model to predict the future rating of roads and inform its budget allocations. DOT engaged a Capstone team in supporting efforts to maximize efficiency through the measurement of relevant rating variables. To quantify the influence of different variables on street ratings, the team chose segment ID-the unique number for each street-as the smallest analyzing unit and regressed segment ID on related variables, including defects, street cut permits, past reconstruction activities, traffic route, and borough. The team established data rules, cleaned the dataset, built a regression model using STATA, and interpreted the results of the analysis. The team's research culminated in the creation of a deterioration model for rating prediction and a set of recommendations to help guide DOT on when and where to execute street reconstruction or resurfacing with available resources.

Capstone Year