NEC used free-to-use satellite images and AI to conduct a demonstration experiment on abnormal inspections of water pipe bridges under the jurisdiction of the Sapporo Water Bureau. As a result, it was found that frequent measurements could be made without the cost of taking pictures, compared to inspection work using drones.

The demonstration experiment was carried out at Toyohira, where the field was provided by the Sapporo Water Bureau. As a result, it was confirmed that the vertical displacement and bridge axis displacement of the water pipe bridge could be measured with an error of about 5 mm, and that the water pipe bridge had no tendency to sag.

Furthermore, it was confirmed that sagging of 3 mm or more, which was artificially generated by simulation, could be detected, demonstrating that it can be used to detect abnormalities in water pipe bridges.

This time, analysis was conducted for eight years from 2016 using free-to-use satellite images observed at 12-day intervals. As a result, frequent measurements could be made without the cost of taking pictures, compared to similar inspection work using drones.

Regular inspections every five years are mandatory for road structures such as bridges that deteriorate over time.
However, with 720,000 bridges nationwide need to be inspected, issues include a shortage of specialist personnel, improving inspection efficiency, and increasing costs due to the need to procure the special equipment needed for inspections and prolonged work.
Visual inspection or an equivalent or better method is also required for water pipe bridges managed by water utilities, but it is difficult to approach long water pipe bridges and conduct visual inspections, which means expensive work costs are incurred using special vehicles.
Recently, the use of drones for inspections has also begun to be considered, but challenges include securing workers skilled in operating them and reducing the associated costs of the work.

※Translating Japanese articles into English with AI