Sund & Bælt Holding A/S has partnered with the Alexandra Institute to market a new solution for inspecting concrete structures. The new solution uses drone photos and artificial intelligence to inspect concrete structures, and it can by itself identify concrete damage when it encounters them.
This new solution builds upon Sund & Bælt’s previous solution for inspection. It can now also be sold to external customers through the new partnership, so that other companies can benefit from it.
Great perspectives for other industries
– We are powering our current solution so that more people can use AI for inspection. It is good for the environment and the green transition when more people can extend the life of their facilities with accurate systematic inspection, explains Bjarne Jørgensen, director of asset management at Sund & Bælt in a post on LinkedIn.
Savings of up to 50 percent
The solution was commissioned in 2022 and shows repeated savings on total inspection costs of up to 50 percent. In addition to the direct savings, it is expected that systematic inspection can extend the life of a bridge by 50-100 years. This not only benefits the economy, but also the climate, as the production of cement emits large amounts of CO2.
Does not make choices or decisions
Business development manager Jesper Rosleff from the Alexandra institute has been the project manager for the development of the new solution. He emphasizes that the artificial intelligence does not make choices or decisions by itself:
– It is a decision support tool for the inspector or engineer. We have trained an algorithm on a lot of images of concrete surfaces and concrete damage so that it can identify damage. But in principle it can be used on all types of surfaces, such as a brick wall, says Jesper Rosleff.
Avoids abseiling on Storebæltsbroen
Especially on bridges, there are big savings to be had. Here you have previously had to send a man up, for example, into the pylons on Storebæltsbroen and rappel down to inspect the concrete. This involves a security risk and is also very expensive:
From eight to two kroner per square meter
– In this way, we go from a price of eight kroner to inspect one square meter of concrete surface to just two kroner per square meter. And with, for example, 360,000 square meters of concrete surface on the Storebælt, it will definitely come to something, says Jesper Rosleff.
According to Jesper Rosleff, a fixed price has not yet been set for the solution, as the algorithm has to be adapted and trained on the individual case:
The drone is the most expensive
– But it is not a particularly expensive solution. The drone itself is the most expensive thing, says Jesper Rosleff, who is clearly impressed by the development within artificial intelligence in the construction industry: – So, if I were 20 years old and an entrepreneur, I know what I would bet on. An insane amount is happening in that area, he concludes.
Facts about the new inspection solution:
- The solution forms an integrated 3D model from thousands of drone photos. This facilitates communication with the contractors who are to repair the identified damage.
- The solution allows for frequent systematic inspection. This makes it possible to follow the development of the individual damages over time, which makes it possible to avoid “over-maintenance”.
- The new method can structure and annotate unstructured images through 3D reconstruction. From the reconstruction, you can precisely determine the position of the drone that takes the pictures and the orientation in relation to the bridge pillar for each individual picture. Thus, it is only necessary to identify each damage once.
Source: Alexandra Instituttet
