Link to Report: Coming Soon
Background :
Bridge infrastructure serves as a critical lifeline for transportation, emergency response, and economic continuity. In hazard-prone regions such as Florida, bridges face escalating risks from floods, hurricanes, and wildfires that can rapidly disrupt traffic flow and delay emergency operations. Existing bridge management systems primarily focus on long-term planning and condition assessment, offering limited capability for real-time decision-making during evolving hazard events. This project aims to develop a real-time decision support framework that enables dynamic management of bridge infrastructure under active hazard conditions. The proposed framework will integrate real-time hazard forecasts, sensor-based condition monitoring, and infrastructure performance data to guide rapid, data-driven decisions. Using advanced analytics and scenario modeling, the system will support time-sensitive operational actions such as rerouting, temporary reinforcement, and emergency closures. A visual decision-support interface will convey hazard progression, bridge condition, and recommended response strategies to transportation agencies and emergency managers in an intuitive, spatially enabled format. Building upon prior work at Florida A&M University on the IntelliViz prioritization platform, this research extends the concept from long-term resilience planning to operational support. A regional case study in Florida will demonstrate the practical implementation of the framework and its benefits for improving coordination, minimizing downtime, and enhancing public safety during flood and hurricane events. By integrating real-time data streams with predictive modeling and visualization tools, the project will bridge the gap between static risk assessment and dynamic hazard response, providing a scalable and implementable framework for strengthening transportation resilience and supporting informed, timely decisions during extreme events.
Objectives :
This project will develop a real-time, multi-hazard decision support framework by integrating predictive hazard models, sensor-based bridge monitoring, and geospatial analytics. Real-time data on hazards will be combined with bridge condition and traffic information to evaluate evolving risks and network accessibility. Machine learning and scenario analysis will be used to identify optimal response actions such as rerouting, reinforcement, or closure. The framework will be implemented within a visual decision-support platform and demonstrated through a regional case study in Florida, showcasing its scalability and operational utility for transportation agencies. This research project has the following four objectives: (1) Develop pipelines for integrating real-time hazard forecasts, traffic conditions, bridge inventory, and operational resource data into a unified decision-support environment. (2) Create a structured database of preemptive and short-term response actions tailored to hurricanes, floods, and wildfires. (3) Develop algorithms for recommending operational interventions based on evolving conditions, infrastructure function, and resource constraints. (4) Illustrate the framework using a case study in Duval County, Florida, providing insights into system performance, applicability, and potential for regional scalability.
Scope :
Task 1: Real-Time Data Integration
This task will establish a centralized data infrastructure to support real-time situational awareness and decision-making. The platform will ingest continuous data streams from multiple sources. Hazard forecasts will include hurricane track and wind field predictions from NOAA’s National Hurricane Center [13], flood inundation and streamflow forecasts from the National Water Model and USGS [14], and wildfire detections and spread estimates from NASA FIRMS and CAL FIRE [15]. Transportation network status will be sourced from FL511, including closures, detour alerts, and traffic flow conditions. Bridge inventory and inspection data will be drawn from the National Bridge Inventory and state-level databases. Crew locations and emergency resources will be integrated through agency input modules. All incoming data will be harmonized and geospatially aligned in near real time, forming the informational backbone of the framework.
Task 2: Response Identification and Planning
This task will develop a structured catalog of feasible operational actions that can be implemented before or during a hazard event to mitigate damage and protect bridges. The primary objective is to define, classify, and organize a set of actionable interventions based on hazard type, bridge characteristics, and available response time. This database will include measures such as temporary shoring, water diversion barriers, fiber-reinforced polymer (FRP) wrapping, modular steel supports, preemptive load restrictions, vegetation clearance, fire-retardant surface treatments, and early detour configuration. Each action will be characterized by implementation time, required resources, level of protection, and compatibility with bridge type and environmental constraints. The framework will associate actions with specific hazard scenarios—such as surge flooding from hurricanes, fluvial flooding from inland rainfall, or advancing wildfire fronts—based on predicted timelines and asset exposure. The resulting action database will be structured for retrieval by the decision-support system and will inform which interventions are feasible at different points along a hazard timeline. This task creates the foundational response options from which Task 3 will evaluate and sequence during a real-time deployment.
Task 3: Decision-Support
This task will develop the core algorithms that convert data and planning into actionable decision recommendations. The platform will utilize exposure modeling to calculate the risk of structural compromise, access disruption, or loss of critical corridors. Multi-criteria prioritization models will consider hazard intensity, the role of infrastructure in the network, redundancy availability, and the timing of impact. Optimization routines will be used to sequence responses based on resource constraints and performance considerations. Machine learning techniques will be explored to classify likely bridge performance under different hazard paths based on past behavior and structural attributes. Scenario simulation will be incorporated to evaluate the outcomes and recommend optimal strategies.
Task 4: Case Study
This task will implement the full framework in a case study focused on Duval County, Florida. The region is a suitable testbed due to its coastal exposure, complex network, and history of recurring natural hazards. The platform will be used to simulate past hazard scenarios and test how real-time decision support could have improved response outcomes. Evaluation metrics will include decision lead time, network performance, and response efficiency. Comparisons will be made between baseline agency response workflows and the framework-assisted scenarios to quantify gains in performance. The application will also support refinement of the platform’s usability and scalability.
Research Team :
Principal Investigator: Neetesh Sharma
Co-Principal Investigator: Eren Erman Ozguven, John Sobanjo