New AI tech allows Florida researchers to monitor the health of buildings
Well-made buildings are said to have "good bones." But if a building or a bridge had broken bones, how would an inspector know?
Doctors use X-rays for patients, and soon local scientists are hoping to put similar monitoring technology in the hands of engineers. Researchers at the University of Central Florida are developing virtual reality and artificial intelligence tools to better monitor the health of buildings and bridges
In 2019, theU.S. ranked 13th on the World Economic Forum for its aging infrastructure. In 2021, the American Society of Civil Engineers gave America's infrastructure a C- and called out a need for more innovative technologies to better monitor and repair the country’s buildings, bridges, and roads.
UCF professor, Necati Catbas, is hoping to address that need with the creation of four different technologies. He and his team of UCF students and postdocs are hoping their tools will allow engineers to check up on buildings the same way a doctor would check on a patient.
“In a way, you're looking at a patient versus you're looking at a patient, and also you're using MRI or X-ray to really understand what's going on,” said Catbas, a Lockheed Martin St. Laurent professor.
"Computer vision" is one such technology UCF researchers have developed to see cracks in infrastructure that in-person inspectors might miss. Using a headset connected to sensors built into a structure, users can see the vibration deformation, and movement of support beams inside a structure. Using mixed reality, users can interact with cracks they spot and use predictive tools to see how they could develop.
Computer vision is made for the visual inspection of structural health, which is practical for inspectors since it doesn’t require access to the structures in question.
"The state of inspection right now is based on visual inspections," Catbas said. "The expertise and know-how of the engineer or inspector is very critical. And that accumulates over the years, but they also need complementary technologies."
Another tool, the generative adversarial network, would allow users to predict how newer structures may crack or shift over a set period of time based on archival data from an older, yet similar structure.
“We are generating new data from the existing data like we are creating synthetic data, and based on the algorithm and methods we can create, and see how the structure is going to look when it has some damage," Catbas said.
The UCF team has also developed an Immersive visualization system that uses virtual reality and augmented reality to conduct “virtual visits” of a building or bridge, from afar. A computer-simulated environment of the real world is generated and overlayed with AR details giving users the structure's status in realtime.
"It's almost like you're having a virtual tour on the bridge," Catbas said. "These are tools to provide more flexibility to have access to the bridge and to have access to the data."
Lastly, the collective intelligence framework technology uses AI to speed up the inspection processes. An inspector uses a headset or a handheld device to scan a damaged area and analyze it in real time. The inspector is spared from performing manual measurements and has access to the building’s condition.
"The ultimate goal here is to effectively manage the data that we are collecting and understand the complex data domains," Catbas said.
Catbas also said these smart structure technologies are ready to be adopted into engineering standards but must reviewed first by many committees throughout the country before they can be applied to everyday engineering and inspection.
Catbas sees the tech as becoming a vital component of America's infrastructure.
"We can utilize these technologies, not only for a particular bridge or bridge assessment, but also for extreme events like hurricanes, floods, earthquakes, and really help people recover from these damaging events," he said. "We can find the critical links in our communities, on the roads, and in buildings. We can find the ones that we need to pay more attention to, work, prepare, and make them more resilient."
Copyright 2023 WMFE.