Automating Concrete Infrastructure Predictive Maintenance

image provided by pixabay

This post is also available in: עברית (Hebrew)

One of the most critical aspects of civil engineering is determining the exact condition of cracks in concrete infrastructure, and AI and other technological advances can ease the task for inspection engineers.

Hugo Nick, an EPFL Master’s student in civil engineering, asked himself how to do so most effectively, and dedicated his Master’s project to solving this problem.

Enrique Corres Sojo, a doctoral assistant at IBETON and one of Nick’s project supervisors, explains: “It’s normal for cracks to appear in reinforced concrete, and these cracks can open or close depending on the type of load they’re subject to. The tricky part is knowing whether a crack is potentially dangerous and if there are any associated risks.”

According to Techxplore, engineers currently inspect structures just by looking at them with the naked eye and simple measurement tools, but such methods can be inaccurate, and some hard-to-reach areas end up not being inspected. That’s why automated crack detection methods have been developed and are being adopted at a rapid pace.

Nick has since graduated and tested two of these new automated methods, assessing the strengths and weaknesses of each.

The first method is called digital image correlation- it is mainly performed in the lab and is known for being extremely accurate. It involves building concrete replicas of structures and placing loads on them to artificially create cracks. The engineers then take digital images during the entire process and run them through special software, which analyzes them and generates displacement fields and deformation fields for the cracks, giving engineers a clear indication of how they’re opening.

The second method is called the finite-segment edge and full-edge approach- it was developed at EPFL’s Earthquake Engineering and Structural Dynamics Laboratory and is still in the experimental stage. It requires inspectors to take a picture of a crack they find on-site, and then a detection algorithm analyzes the picture using artificial intelligence. Nick further explains: “The algorithm is actually a neural network that’s been trained on thousands of pictures and can predict crack detection. Inspectors are starting to test this method out in the field, and it has several advantages.”

Nick researched the capabilities of the second method and found that to obtain accurate readings of a 0.3 mm crack, the camera needs to be 35 cm away from it. This confirms the method is easy to use, can be employed with a smartphone or a drone, and therefore is suitable for hard-to-reach areas.

That said, the algorithm’s accuracy still has a way to go, and digital image correlation is currently more reliable (especially for very small cracks), so both methods should be used going forward to help inspection engineers collect more complete data on the general condition of reinforced concrete structures.