Developing a Low-Cost Pavement Management System for the Borough of Collingswood

Status:

Funding Agency: Borough of Collingswood

Abstract: 

This project designed a cost-effective, AI-driven pavement management system for Collingswood, NJ. The team used smartphone-based imaging and convolutional neural networks (CNNs) to detect cracks and potholes with 82% accuracy. A regression-based PCI model provided automated condition assessments and treatment prioritization. The system created a full roadway inventory and recommended evidence-based maintenance strategies. This project demonstrated how AI and affordable tools can improve asset management for small municipalities.

Principal Investigators: Dr. Yusuf Mehta, PhD., Dr. Ayman Ali, PhD.

Link to Publications: