Using Electric Vehicle Onboard Data for Pavement Quality Assessment and Management

Status: Complete
Report Date: 01/20/2026

Summary:

Onboard data from electric vehicles can effectively assess pavement conditions for continuous infrastructure monitoring. This methodology leverages vast amounts of low-cost, low-quality data into valuable pavement condition information to achieve more comprehensive coverage of Minnesota’s entire state highway system. Investigators evaluated numerous machine learning models by comparing their predictive output to pavement assessments conducted by MnDOT’s Pathways van—the current method used to assess pavement condition. The best-performing model predicted pavement roughness with approximately 94% accuracy. MnDOT has recently contracted with a firm that collects similar network-level data to take advantage of the benefits of onboard vehicle data.

Final Deliverables:

Using Electric Vehicle Onboard Data for Pavement Quality Assessment and Management (Report #2026-05)
Using Onboard Vehicle Data to Assess Pavement Quality (Research Summary)

Related Materials:

Using Onboard Vehicle Data to Assess Pavement Quality (Blog Post)

Related Research:

Project Personnel:

Principal Investigator: Mihai Marasteanu

Co-Principal Investigator: Qizhi He, Raphael Stern, Mugurel Turos, Shaghayegh Nouhi

Technical Liaison: Curt Turgeon

Project Coordinator: Barbara Fraley