Drones and AI can reduce risks in corroded bridges

Today, corroded bridges pose significant safety risks, but advances in drone and AI technology are enabling more efficient and accurate detection and maintenance. For this purpose, Japanese technology companies NTT Corporation and NTT e-Drone Technology Corporation have collaborated to develop a method for detecting corroded steel on bridges using drones and AI-based image recognition. The approach reported can not be compared to traditional visual inspections since it enables early detection and accelerates the necessary repairs for steel bridges. Thus, it can be considered that bridge safety is improved, and maintenance cost is reduced.

Steel bridge maintenance is challenging due to corrosion occurring in areas hard to be reached. Traditional inspection methods are insufficient since in most cases rely on visual assessments. In fact, they can detect corrosion, but they cannot measure its depth. To address the latter, NTT has developed a solution using drones that are equipped with cameras and advanced AI image recognition. The drones take images of the bridges, and the AI analyzes these images to detect damage. To assess the damage and check AI’s performance, experts always review the AI’s findings.

The approach proposed, removes the need for expensive scaffolding and manual ultrasound probes, while also reducing the potential for human error. It cannot be omitted that in traditional inspections, results depend heavily on the engineer's experience. In comparison with traditional approaches, in this new system, assessors only need to make a final judgment on the damage, improving the efficiency and precision of inspections.

The two NTT companies tested the technology on a case-study steel bridge located in Kumagaya City, Japan. Initially, the drones autonomously captured detailed images. The images were then analyzed by AI to identify corrosion. This automated process can be seen as cost-effective while it can be used more frequently than a conventional method. The companies aim to enhance the technology’s accuracy and improve drone efficiency. Their plans include the commercial deployment by 2025 while they believe that the system could be applied to other infrastructure, such as steel towers or road crash barriers.