PI: Sean Qian
Co-PI(s): N/A
University: Carnegie Mellon University
Industry partner: RoadBotics Inc.
Each year, the City of Pittsburgh resurfaces 50-90 miles of urban roads, and the roads deteriorate over the years, adding an additional cost to both the city and residents. Pittsburgh is not alone. Deteriorating pavements and the ability to effectively assess and retrofitting those are common issues among all cities and communities in the U.S. The ultimate goal of this research project is to establish a pavement management system that monitors and predicts pavement conditions for each street, estimates paving expenditures, and ultimately identifies optimal paving schedules for the city. Pavement deterioration is affected by various factors, depending on weather conditions, pavement characteristics, and how the pavement has been used in the past. Currently, the city of Pittsburgh (and most cities too) only uses age and functional class to forecast deterioration and determine paving plans, while other factors such as weather and traffic loadings play a vital role in the deterioration. For example, bus stops are likely to cause pavement distress and truck loadings can cause pavement cracking, and their effects are substantially more pronounced than a standard passenger car. Snowing also affects the pavement as a result of anti-ice products sprayed on the roads, let alone the notorious freezing and thawing cycles that accelerate pavement deterioration. The Mobility Data Analytics Center (MAC) has archived many of the datasets including road characteristics, traffic counts and speed by vehicle classes, land-use, weather conditions, and right-of-way permits. RoadBotics Inc. a PA-based technologies firm, has worked with many domestic cities to measure pavement ratings leveraging their AI technologies. Those datasets altogether have great potential to enhance traditional pavement deterioration models. We propose to develop a data-driven predictive deterioration model to better understand the effects of various factors on pavement conditions, followed by a pavement management model that generates optimal maintenance plans for communities.