UT Dallas AI Team Helps Japanese Highway Company Build City Road Repair System
Researchers at the University of Texas at Dallas have partnered with NEXCO-Central, a Japan-based expressway operator, to develop software that helps cities decide which roads to fix first when budgets are tight.
The system adds a prioritization layer to NEXCO-Central's existing technology, which uses mobile cameras and artificial intelligence to assess pavement conditions across entire road networks. The new tool automates decisions that city managers typically make by hand-weighing repair urgency, costs, and competing needs.
"The new system emulates the mind of a city manager who has to decide the priority for fixing various road segments," said Dr. Gopal Gupta, director of the UT Dallas Center for Applied AI and Machine Learning.
How It Works
The software includes a scoring system that ranks road segments based on condition data and financial constraints. It also explains the reasoning behind each recommendation, giving city officials visibility into how the algorithm reached its conclusions.
NEXCO-Central serves clients mainly in North Texas through its subsidiary NEXCO Highway Solutions of America Inc. The company approached UT Dallas to move beyond simple condition assessment into complex decision-making.
"Our technology aims to optimize the complex decision-making to determine which roads are most in need of repairs, the predicted financial investment and prioritizing who gets the money and when," said Koshiro Mori, a developer at NEXCO-Central.
The Collaboration Model
The partnership operated under an Intellectual Property Assignment Sponsored Research Agreement, which lets companies retain ownership of technology developed through the collaboration. Computer science doctoral students Abhiramon Rajasekharan and Keegan Kimbrell contributed to the project alongside Gupta.
Mori said NEXCO-Central found UT Dallas after searching online for institutions with expertise in AI and machine learning. "We saw a collaboration opportunity, and we're very happy with how the team has handled this project," he said.
The Center for Applied AI and Machine Learning has worked with companies globally on similar projects, with most resulting in the partner company owning the intellectual property. Recent collaborations include work with energy company Vistra and healthcare startup CorroHealth Inc.
For science and research professionals interested in applying AI to industry problems, explore AI for Science & Research or consider AI Research Courses that cover machine learning applications in real-world settings.
Your membership also unlocks:
