MPC Research Reports |
Title: | Mobile Phone-Based Artificial Intelligence Development for Maintenance Asset Management |
Authors: | Biao Kuang and Jianli Chen |
University: | University of Utah |
Publication Date: | Jul 2024 |
Report #: | MPC-24-533 |
Project #: | MPC-668 |
Transportation asset management requires timely information collection to inform relevant maintenance practices. Traditional data collection methods often necessitate manual operation or the use of specialized equipment, e.g., light detection and ranging (LiDAR), which can be labor-intensive and costly to implement. With advancements in computing techniques, artificial intelligence (AI) has emerged as a powerful tool for automatically detecting objects in images and videos. Therefore, this project aims to develop an implementable AI package that streamlines the inspection of transportation assets through automated processes. Specifically, a smartphone was mounted on the vehicle's front windshield to record videos of transportation assets on both highways and local roads in Utah. These videos were then converted and processed into labeled images, which served as training and test datasets for the AI algorithms. Based on a deep learning framework, i.e., You Only Look Once (YOLO), we developed accurate and efficient AI algorithms to automatically detect and identify transportation assets, which include pavement marking issues, traffic signs, litter and trash, and steel guardrails and concrete barriers. The developed AI models demonstrate good performance in identifying targeted objects, achieving over 85% accuracy. The AI package developed in this project is expected to enable timely and efficient information collection for transportation assets, thereby improving road safety.
Kuang, Biao, and Jianli Chen. Mobile Phone-Based Artificial Intelligence Development for Maintenance Asset Management, MPC-24-533. North Dakota State University - Upper Great Plains Transportation Institute, Fargo: Mountain-Plains Consortium, 2024.