Mobile Phone-Based Artificial Intelligence Development for Maintenance Asset Management [Research Brief]

Chen, Jianli; Kuang, Biao · 2024 · ROSA P / Mountain-Plains Consortium

archive: archived pipeline: cataloged verified

Get this paper ↗ (full text — opens at the source; we link to it, we don't host it)

Summary

This research brief addresses the need for low-cost, automated alternatives to traditional transportation asset assessment methods, which are often labor-intensive or prohibitively expensive for periodic inspections. Timely information collection is critical for state departments of transportation to guide daily maintenance practices. The study aims to develop a mobile phone-based artificial intelligence (AI) system capable of identifying and assessing transportation assets efficiently, thereby enabling more frequent inspections and improved road safety. The research was conducted through four specific tasks. First, a comprehensive literature review examined existing technologies and emerging AI practices in transportation asset collection. Second, data was collected using a mobile phone mounted on a vehicle driving on Utah highways and streets, recording videos to pre-evaluate AI capabilities for asset identification and condition assessment. Third, multiple AI models were developed using self-collected images to inspect pavement marking conditions, identify traffic signs, and detect common road litter. Fourth, a prototype AI model was developed specifically to identify concrete barriers and steel guardrails. The performance of each model was evaluated through iterative training and tuning to ensure robustness. The findings demonstrate the successful development of three primary AI models for automatic detection. The pavement marking model detected faded white and yellow markings. The traffic sign model identified regulatory, speed-related, warning, and guide signs. The litter and trash model detected white litter, black litter, dirt, and leaves on the roadside. Additionally, the prototype model successfully identified steel guardrails and concrete barriers. The results indicate that the developed AI models achieved good performance, with over 85% accuracy in transportation asset identification. The mobile phone-based AI package provides an accurate, efficient, and automated approach for collecting and analyzing transportation asset data. The significance of this work lies in its potential to transform maintenance asset management by offering a scalable, low-cost solution for data collection. By enabling more frequent and accurate inspections, road agencies can target timely maintenance activities to extend the life of road assets. This automated approach ultimately improves safety for road users by ensuring that asset conditions are monitored effectively without the constraints of traditional manual or costly device-based methods. The study highlights the viability of using consumer-grade mobile technology combined with AI algorithms to support critical infrastructure management decisions.

Key finding

Mobile phone-based AI models achieved over 85% accuracy in identifying transportation assets such as pavement markings, traffic signs, litter, barriers, and guardrails.

Methodology

field_study

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed. Discovered via bulk_ingest_rosap on 2026-05-23 (6 acquisition events logged).

Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• ehmi external hmi