New bridge-inspection software contributes to bridge security
Inspecting a bridge for hairline cracks, flaking concrete, and rust has been a manual process — inspectors have always examined bridges for visible damage directly on site; German researchers develop software which allows digital inspection of bridges and other structures
The changing effects of weather and temperature, road salt, and traffic all take their toll on bridges, causing damage such as hairline cracks, flaking concrete, and rust. In extreme cases, they contribute to the collapse of a bridge, as happened in Minneapolis last year. Until now, inspecting a bridge has been a manual process — inspectors have always examined bridges for visible damage directly on site.
A new image processing software developed by researchers at the Fraunhofer Institute for Industrial Mathematics in Kaiserslautern and scientists from the Italian company Infracom is about to change all this. The new software they have written can examine a photograph of a bridge for irregularities and then mark the respective area on the photo to alert an inspector to any problems. Building the software was not easy, as bridges differ in terms of their shape, construction material, and surface structure. The color of bridges, for example, depends on the material, the dirt or fouling, and the degree of humidity.
The software developers thus had to build a database of images featuring many different types of bridges that had experienced many different sorts of damage, while categorizing the type of damage with the visual image itself. The results of the development effort have paid off. Now, when a new photograph of a bridge the characteristics of which are are unknown is presented to the system, that image is compared with the images already stored in the database. The software can then determine whether there are any irregularities. If there are, it marks the respective area on the photograph, allowing a bridge inspector to decide how serious the damage is.
Engineers have already been using the software over the past six months to inspect bridges in Italy. “Im Gegensatz zum Menschen übersieht das Tool keine Auffälligkeiten — auch kleinste Schäden werden identifiziert und markiert” (“Unlike a human, the tool doesn’t miss any abnormalities - even minor damage is identified and signaled”), said ITWM scientist Markus Rauhut.
