The earlier problems are identified, the faster and usually more cost-effectively they can be remedied. What in many cases primarily brings economic benefits can also be safety-relevant in other places. For example, when it comes to detecting cracks and other damage on bridge piers, dams or even runways. A project at Dübendorf Airport in Switzerland is now testing how this could be achieved particularly efficiently in the future using drone photos and AI.
The use of drones to detect damage to structures or roads in order to facilitate the generation of image data or to optimize the perspective is not new. AI-supported data evaluation is also no longer a dream of the future. However, artificial intelligence needs to be trained to perform certain tasks on its own. This is because damage is structure-specific; cracks in a bridge pier can be very different from those on a runway. For the best possible results, data from the intended use case must therefore be incorporated into the creation of new models - a time-consuming training process...
The full article is available on Drones.