AI-Enabled Energy-Adaptive Digital Twins for Sustainable Bridge Health Monitoring
Presented By: Hadi Salehi
Affiliation: Louisiana Tech University
Description: Digital twins are emerging as powerful tools for bridge structural health monitoring (SHM), enabling real-time synchronization between the physical structure and its virtual representation. However, their reliability and autonomy remain limited by the continuous energy demands of wireless sensor networks. Most existing frameworks assume stable power and data availability, overlooking the practical constraints of long-term field deployment where sensing and communication must operate intermittently. A critical knowledge gap exists in developing energy-adaptive digital twins that incorporate the physics of hybrid self-powered sensors and manage energy availability as part of the monitoring strategy. This study introduces an AI-enabled, energy-adaptive digital twin that integrates finite element (FE) simulations, hybrid self-powered piezoelectric sensors, and intelligent model updating. Each sensor harvests mechanical energy from bridge vibrations to power the sensing process while relying on limited stored energy for data transmission to gateways. The digital twin continuously co-simulates both the structural and energy domains, using AI-based models trained on FE data to predict strain fields, harvested energy, and power usage across the sensor network. These forecasts allow the system to dynamically determine when and where to collect data, optimizing information gain under energy and bandwidth constraints. By embedding energy-awareness and AI-driven adaptivity into its architecture, the framework transforms the digital twin into a self-regulating, resilient, and sustainable SHM system. Numerical studies evaluate its capacity for stiffness degradation and long-term autonomy. This study aims to advance the next generation of intelligent, energy-conscious digital twins for durable bridge infrastructure management.
Monitoring-Informed Structural Analysis of an Existing Bridge using Displacement Data
Presented By: Mohamed ElBatanouny
Affiliation: Wiss, Janney, Elstner Associates, Inc.
Description: Bridge construction and widening projects often require installing new foundations adjacent to existing structures, which can induce vibrations and soil movement. In loose sandy soils, these activities may cause settlement or lateral displacement of existing bridge foundations, posing risks to structural integrity and public safety. While some bridges remain unaffected, others exhibit measurable movement that demands close monitoring.
The IA-9 / WI-82 Mississippi River Bridge in Lansing, Iowa is undergoing a replacement project where potential movement of the existing bridge was anticipated. To mitigate risk, a comprehensive monitoring system, comprising tilt sensors, GPS displacement units, and survey control points, was deployed to track real-time bridge behavior. During construction, recorded displacements exceeded project-defined limits, triggering further evaluation. Monitoring data were integrated into a 3D finite element model of the existing bridge to perform structural analyses and confirm its continued safety for traffic. Periodic updates of displacement data allowed iterative model refinement, ensuring ongoing public safety throughout the project. This case study demonstrates the critical role of structural health monitoring in supporting model updating and advancing digital twin applications for infrastructure management.
Digital Twin for Predictive Maintenance: Implementation at the Köhlbrand Bridge, Hamburg / Germany
Presented By: Horst Trattnig
Affiliation: Vallen Systeme Gmbh
Description: Maintaining the functionality and availability of our transportation networks is a growing challenge, requiring more digitally supported maintenance. To optimize this process and implement predictive analytics, we can leverage the power of digital twins. By moving toward the predictive maintenance approach, the validation of simulated data, and real-world Acoustic Emission sensor data, are key to close the loop for the longevity of critical assets. This presentation shows a digital twin implementation at an infrastructure. It provides insights into the implementation of a digital twin for the Köhlbrand Bridge in Hamburg / Germany, demonstrating how simulated and real-world Acoustic Emission data are processed to create a digital physical model.
Lessons Learned from Using Field-Verified Digital Twins for 35+ Years
Presented By: Marisol Tsui Chang
Affiliation: Bridge Diagnostics, Inc.
Description: Digital twins are becoming a practical tool for infrastructure asset management, offering a way to combine advanced modeling tools with field-verified data to better understand how structures perform over time. This presentation shares lessons from more than 30 years of applying digital twin methods to assess and manage concrete infrastructure. It highlights how field-verified investigations, paired with data-driven decision tools, can support more accurate evaluations and long-term planning. Case studies include condition assessments of in-service concrete structures using a mix of nondestructive testing, sensor data, and modeling techniques. These examples show how digital twins can help identify deterioration early, reduce uncertainty in maintenance decisions, and improve overall asset reliability. This presentation will also discuss common implementation challenges and offers practical recommendations for engineers looking to integrate digital twin approaches into their workflows.