Bridge inspections are time-consuming, inefficient and often subjective
processes that rely heavily on human judgment, leading to overlooked or
underestimated structural problems or damage. A new process developed by
Arvin
Ebrahimkhanlou, PhD, assistant professor of civil,
architectural and environmental engineering at Drexel University, aims to
address this issue.
Using graph theory algorithms, a system converts images of concrete cracks
into mathematical graph representations with features like assigned values
for the intersecting cracks. Machine learning algorithms then correlate
these graphed crack features to different damage levels based on the crack
sample data.
“Creating a mathematical representation of cracking patterns is a novel idea
and the key contribution of our recent paper,” Ebrahimkhanlou said. “We find
this to be a highly effective way to quantify changes in the patterns of
cracking, which enables us to connect the visual appearance of a crack to
the level of structural damage in a way that is quantifiable and can be
consistently repeated regardless of who is doing the inspection.”
In tests, the AI correctly assessed over 90% of cracked samples’ damage
levels. This rapid quantification method could identify problems even before
human inspectors notice them visually. By leveraging AI to impose order on
chaotic visual features, this innovation aims to pinpoint infrastructure
problems before catastrophic failures occur.