Four new inventions for monitoring the structural health of the nation’s aging buildings, roadways and bridges, among other structures, are proposed by researchers from the University of Central Florida (UCF).

According to the researchers, traditional methods for monitoring such structures usually involves visual inspection, which can result in costly and time-consuming closures and potentially dangerous conditions for those performing the on-site inspections. As such the researchers are turning to assorted solutions that use artificial intelligence (AI), virtual reality (VR), augmented reality (AR) and computer vision to monitor such structures.

Source: University of Central Florida

The first proposed solution relies on computer vision wherein a comprehensive structural health monitoring system lets inspectors view and assess the load-worthiness and serviceability of structures via cameras stationed around the structure in question, collecting image and location data. As the cameras monitor the site, the computer vision software analyzes the data and offers users of the system safety assessments, including information about structural changes, structural weaknesses and structural damage.

The second proposed solution involves VR and AR for analyzing structures through so-called "virtual visits," wherein VR offers an entirely computer-simulated environment as AR produces content onto views of actual real-world environments, virtually bringing inspectors to disaster areas so that they can virtually inspect bridges and buildings, for instance.

Similar to the first solution suggested by the researchers, this second solution relies on sensors and cameras. However, this technology, according to the team, also relies on robots, drones, lidar scanners and infrared thermography cameras to offer a real-time view of a site and the ability to interact and communicate with people both onsite and at different locations.

Meanwhile, solution number three relies on a combination of AI and mixed reality to expedite inspections. The researchers explain that this invention enables an inspector to stand outside a damaged building or near a damaged bridge, wearing a headset or using a hand-held device featuring AI and mixed reality technology.

For instance, the inspector might use one of these items to scan the damaged areas, as the system analyzes the structure in real-time, eventually calculating or assessing the structure's condition, thus expediting the inspection process.

Finally, solution number four employs a generative adversarial network (GAN) that relies on AI for predicting damage and thereby minimizing the need for data collection from several structures, with inspectors collecting data from just a few structures, instead of installing sensors and devices on all structures.

"There is not enough data from damaged areas to train detection models," the researchers explained. "Yet, machine learning (ML) and deep learning (DL) algorithms used with AI yield better, more accurate output using big data sets. As a solution to the data scarcity in civil structural health monitoring applications, the invention takes data collected from structures. It uses model variants of the GAN architecture to generate large, accurate synthetic data samples to train damage diagnostics systems."

The UCF researchers added that AI could better inform inspectors about what's happening with other similar structures, making predictions about structure changes before they occur, and thus decide on an effective response.

The researchers suggest that the four solutions could potentially be used in combination or independently.