Project Details

Abstract

One key part of pavement management is to assess the road condition and identify pavement distresses such as cracks and potholes. These road distresses, if not identified and repaired timely, could compromise road safety, cause expensive damage claims, and also lead to more expensive later repairs. To assess pavement condition, pavement condition data need to be collected first. However, traditional pavement data collection still relies on manual or specialized vehicles equipped with expensive sensors and requires personnel driving along each road in the road networks. Therefore, traditional road inspection methods are often costly, labor-intensive, and sporadic with limited coverage, leading to delayed maintenance and compromised safety. Recent advancements in machine learning (ML) and the proliferation of electric vehicles (EVs) equipped with various sensors offer a promising avenue for revolutionizing road condition assessment practices. This project will establish a framework for collecting and processing rideshare crowdsourcing EV data, and develop machine learning algorithms that uses data from EVs to automatically assess road conditions and identify road damages such as cracks and potholes. The project has the potential to offer a more efficient, cost-effective, and real-time approach to road condition monitoring over large road networks and provide critical information for timely maintenance.

Project Word Files