Concrete and Steel Defect
Segmentation for IHI Corporation

ABOUT THE CLIENT

With over 30,000 employees and over 8 billion USD in revenue, IHI Corporation is a comprehensive heavy-industry manufacturer working to create value for customers in four main areas – Resource, Energy, and Environment; Social Infrastructure and Offshore Facilities; Industrial Systems and General-purpose Machinery; and Aero Engine, Space and Defense. IHI’s history extends to establishing Japan’s first modern shipbuilding facility in 1853. Since then, the company has played a crucial role in Japan’s modernization.

46% retention rate from referrals vs 14% from career sites

Hiring process made
30% faster

Solution

Current inspection methods for large critical infrastructures, such as bridges and dams, involve climbers on ropes or scaffolding and are slow, expensive, and dangerous. In partnership with a drone company, we could use drones to fly over concrete structures and capture thousands of images with no risk to people or equipment. Then our proprietary AI models trained on similar infrastructure will process the images to segment the defects such as cracking, crazing, blistering, delamination, efflorescence, spalling, corrosion, and exposed rebar. This pipeline helps IHI identify defective areas in hours without spending a lot of money and months of labor.

FRAMEWORKS USED