Sarcouncil Journal of Applied Sciences Aims & Scope

Abstract: The aging condition of U.S. bridge and highway infrastructure, combined with rising maintenance backlogs and the economic consequences of unplanned failures, has intensified the need for proactive, data driven maintenance strategies. The objective of this paper is to address the existing gap by delivering an extensive analysis of machine learning algorithms for predictive maintenance within U.S. bridge and highway infrastructure in recent years, emphasizing their capacity to mitigate economic disruptions. This review synthesizes peer reviewed studies and agency reports from 2020–2026 to evaluate the application of machine learning for predictive maintenance aimed at minimizing economic disruptions. Using a qualitative review methodology, the study examines machine learning models used for condition prediction, deterioration modeling, anomaly detection, and maintenance optimization across bridges, pavements, and highway networks. Findings show that tree based ensemble models, particularly random forests and gradient boosting, consistently deliver the strongest predictive performance for bridge and pavement condition ratings, while deep learning and multimodal vision models excel in automated distress detection from imagery. Emerging research integrates these predictions with optimization routines to support budget constrained maintenance planning, though most studies remain asset specific and rarely quantify avoided user delay costs or broader economic impacts. Key challenges include data quality limitations, lack of standardized benchmarks, limited uncertainty quantification, and slow institutional adoption. The review highlights the need for integrated, network level frameworks that couple machine learning predictions with traffic assignment, user cost modeling, and multi objective optimization to better align predictive maintenance with economic resilience goals. These insights underscore the potential of machine learning to enhance early stage decision making, extend asset life, and reduce economic losses associated with infrastructure failures.