AI and virtual reality to monitor the structural safety of buildings and bridges

The structural health and safety of a nation's bridges and buildings is crucial for preventing disasters like the Surfside condominium collapse in 2021. To improve the structural health monitoring of various structures, e.g., buildings, bridges, or roads, researchers at the University of Central Florida have introduced four innovative solutions. What is of special interest is that the inventions proposed are based on Artificial Intelligence (AI) and Virtual Reality (VR).

Necati Catbas, a Lockheed Martin St. Laurent Professor in UCF's Department of Civil, Environmental and Construction Engineering, emphasizes the need for structural health monitoring, likening it to human health monitoring. With the proposed technologies, the researchers aim to address challenges that are linked with the infrastructure that suffers the effects of aging. It should be mentioned that traditional methods involve on-site visual inspection and hence are time-consuming and costly, leading to road or bridge closures and potential dangers for personnel. Catbas and his team's inventions aim to mitigate these issues.

One invention utilizes computer vision, extracting meaningful information about structures through image analysis. Thus, the need for direct access to bridges, buildings, or towers is eliminated. This technology allows inspectors to remotely assess structural conditions, anticipating risks and prioritizing investments. Specifically, the invention uses cameras to collect image and location data. A computer vision software continually processes and analyzes this data, providing a safety assessment that includes information about structural changes, weaknesses, or damage.

The second invention involves an immersive visualization system utilizing VR and AR for "virtual visits." This technology enables experts to virtually inspect disaster areas, providing real-time views and facilitating communication with colleagues globally. Damage detection and load-carrying information about a structure is provided using cameras and sensors while other tools, e.g., robots, unmanned aerial vehicles (UAVs) or drones and/or LiDAR scanners, are also employed. The result is always a real-time view with the ability to communicate and interact with people.

Two additional inventions incorporate AI to enhance inspections and structural damage diagnostics. The collective intelligence framework blends human-centric AI with mixed reality, streamlining inspection processes and improving accuracy. In addition, the generative adversarial network (GAN) takes a proactive approach to structural health by using AI to predict damage. In the first case, the system saves the inspector from performing manual measurements since during the assessment, the inspector interacts with the AI adjusting its defect. The system uses all changes to train the AI model and hence to improve its accuracy. In other words, it can combine professional judgment with the AI's power. Concerning the second case, the technology predicts the dynamic response changes before damage conditions occur. The technology also has the ability to simulate scenarios of structures to illustrate how a well-maintained bridge would respond after damage in comparison to the response of a poorly maintained bridge.

Catbas envisions future projects, including a framework for smart and resilient communities to withstand extreme events. This framework aims to improve disaster preparedness, resource allocation, and evacuation planning. The team is also developing a "digital twin" of infrastructure assets, like NASA's use of replicas for spacecraft components, allowing simultaneous data collection and predictive analysis for different structural scenarios.