AI & Climate Change: Protecting Sweden's bridges through real-time data
The increasingly unpredictable climate has made society’s infrastructure more vulnerable than ever. Rainstorms and other extreme weather conditions have had devastating effects on Sweden’s bridges and roads. Artificial intelligence (AI) and advances in the field hold the key to increasing the resilience of these key structures, according to KTH researchers in Sweden.
According to Join Leander, a KTH researcher working on AI-based imagery analysis for infrastructural safety assessment, inspecting bridges manually, and in person is both time-consuming and risks missing critical defects, resulting in further damages and dangers for people and key transport infrastructure. In his words, "this is where AI and modern technology comes in as a game changer".
Towards sustainable infrastructure: AI's role in helping us adapt to climate change
Leander and his team are working to make inspections more effective and accurate by using drones and machine learning. By analysing the images collected by drones, AI algorithms can identify damage and where that could be overlooked in a manual check.
“This technology has the additional advantage of being able to operate quickly, which is particularly important in a crisis situation such as a major storm or flooding,” - adds Leander.
In addition to drones, sensors mounted directly on bridges are also used to collect real-time data. “This is data that can be continuously analysed to give a snapshot of the status of the bridge and this allows us to act proactively instead of reactive,” he explains. Real-time data opens up a world of possibilities - from reduced maintenance needs to establishing a full monitoring cycle, where data on issues and problems is seen, analysed and acted upon by someone almost simultaneously.
Real-time data: saving money, saving time
According to Leander, it is precisely these technological advances like real-time data that could save time and money. “At present, many decisions are taken to demolish and build new bridges based on theoretical life calculations which often turn out to be incorrect. With real-time data, these decisions would be more based on reality".
However, Leander does not hide that if these methods are to become internationally recognised, and accepted standards, further research and testing is needed - urgently.
“We still have some way to go. We need to collect more data from different types of bridges and different climatic conditions to ensure that our technology is reliable, "says Leander.
At the same time, he sees a bright future for AI in this context, not only to address the acute consequences of extreme weather, but also to adapt Sweden’s infrastructure to the changing climate.
“AI has the potential to solve not only the pressing problems we face today, but also the long-term challenges. It is a powerful tool to make our infrastructure more sustainable and resilient".
Figure: Unsplash ©
