Detecting corrosion under coatings using a portable smart spectrometer powered by machine-learning techniques

Efforts for efficient ways to do routine AF fleet maintenance for inspection have been made by utilizng advanced imaging technologies for nondestructive inspection, including visual, ultrasonic, electromagnetic, radiographic, and Long-Wave Infra-Red (LWIR) or thermographic imaging. We suggest that LinkSquare be a unique SWIR solution in providing through-paint inspections without protective coating removal to mitigate the abovementioned problems. LinkSquare’s SWIR technology, coupled with its machine learning techniques, will be a quick and reliable nondestructive inspection method for detecting and characterizing hidden damage under an aerospace coating system. LinkSquare illuminates broader spectrum light that typically penetrates 5 mm below a coating and uses machine learning algorithms to determine if corrosion is present and display the result on the screen in seconds. Each of existing approaches of visual, ultrasonic, electromagnetic, radiographic, and thermographic imaging has its own advantages and disadvantages in terms of measurement sensitivity, resolution, area coverage, ease-of-use, quantitative assessment, data processing requirements, and inspection speed; however, LinkSquare boasts itself in that it can deliver a full package solution – not only the LinkSquare device but the software to actualize the customer needs.