AI and Drones Revolutionize Bridge Maintenance
NTT Corporation, along with NTT e-Drone Technology, has devised a novel way to find and fix rust on steel bridges using drones and AI.
Tackling Corrosion Challenges
Steel bridges often face issues from rust, especially in spots that are tough to reach. Normally, inspectors check for rust by looking at it, but that doesn’t always show how bad it is. NTT shared, “It can detect corrosion and estimate the corrosion depth at the same time by imaging with drones and AI inspection, which has been difficult to do with visual inspection by inspectors.” With drones and AI, it’s easier to spot and measure rust without needing expensive tools or scaffolding.
The Federal Highway Administration points out that finding rust isn’t easy because results depend on who’s checking and the tools used. NTT’s tech uses AI and drones together, making inspections more accurate and efficient. This is key for keeping old bridges strong.
We think making things easier and cheaper with technology is pretty cool.
Testing in Japan
In Kumagaya City, Japan, NTT tried their drone and AI system on a bridge. Drones took pictures, and AI looked at these to find rust and figure out how deep it went. NTT says this method could cut inspection costs and be more precise.
NTT stated, “Improvement of work efficiency and reduced maintenance costs by using an ultrasonic device to measure the amount of loss in the steel cross section of corroded areas using drone imaging and AI inspection.” By making these tasks automatic, keeping bridges in good shape becomes easier and cheaper.
Looking Ahead
NTT wants to use this tech on other things like steel towers and guardrails too. They plan to make AI better and drones more efficient, aiming to have this ready for inspections by 2025.
They hope to help society by cutting down on maintenance costs and making sure important structures are safe. Using what they learned in Kumagaya City, they’ll see how practical this is and think about using it on more types of infrastructure.
