This project continues to advance transportation safety through the deployment and scaling of cutting-edge technologies for near real-time monitoring of infrastructure. Building on the foundational work of Year 1 (inverse problem framework development and vision-based structural health monitoring for bridges) and Year 3 (AI-powered video analytics scaled to urban infrastructure), the Year 4 initiative at Howard University integrates three funded years of research into a deployable, scalable monitoring solution.
The core focus of Year 4 is field deployment of the integrated monitoring platform on an operational bridge in Washington, DC, identified in coordination with the District Department of Transportation (DDOT). In parallel, the project will conduct exploratory trials of drone-based video and imagery for structural inspection, in collaboration with DDOT’s Drones team, to assess whether this complementary sensing approach warrants full integration in a potential Year 5. The work integrates civil engineering, computer vision, machine learning, and applied mathematics for AI-powered structural diagnosis.
The primary deliverable will be a validated, field-deployed monitoring system operating on a selected bridge in Washington, DC, capable of processing incoming video and structural data and providing actionable recommendations for maintenance prioritization. The system integrates the signal-processing, inverse-problem, and machine-learning components developed across the project’s three funded years, deploying AI-based computer vision merged with signal-processing methods, such as feature tracking, optical flow, motion magnification, Enhanced Frequency Domain Decomposition, and Stochastic Subspace Identification, for automated anomaly detection and structural parameter identification from surveillance camera data. The exploratory drone trials will produce a feasibility assessment of aerial visual data for structural monitoring across multiple infrastructure types. Additional deliverables include a documented TDI deployment model suitable for replication by other agencies, an archive of open-source code and labeled video datasets, and supporting technical documentation.
Universities Involved
Howard University
Principal Investigators
Claudia Marin
Expected Research Outcomes & Impacts
The field-deployed monitoring system will enable transportation agencies to transition from reactive maintenance toward proactive, data-driven infrastructure management, advancing infrastructure monitoring by introducing a more efficient and scalable structural health monitoring methodology that enhances damage detection without requiring costly sensor networks. The exploratory drone trials will clarify whether aerial visual data can extend monitoring across multiple infrastructure types.
By delivering a documented and replicable deployment model, an open-source code and dataset archive, and technical documentation, the project supports broader adoption by other agencies. Findings will be disseminated through national and international conferences, including the Transportation Research Board Annual Meeting and American Society of Civil Engineers conferences, and through peer-reviewed journals, while integrating research into Howard University courses and student research to build the next generation of transportation professionals.
Subject Areas
Structural Health Monitoring, Bridge Infrastructure, Computer Vision, Machine Learning, Asset Management
