Automated System for Surface Pitting Analysis on Airplane Jet Engine
Posted: 2018-4-30
Principal Investigator: Dr Stephen O'BRIEN, Aviation Services Research Centre.
"Pitting" is a surface defect found on critical jet engine components of complex shapes and different sizes. It is currently detected by naked-eye inspection. Such task takes up significant man-hours, and embodies errors in accuracy and reliability which limit engine maintenance performance in terms of cost, time, quality, and safety. The newly-developed automated system by Aviation Services Research Centre (ASRC) of Hong Kong Polytechnic University employs robotics, artificial intelligence and deep learning, image processing, 3D scanning, optical metrology, and non-destructive testing technologies to identify and measure "corrosive pitting" such that precise information such as pitting position, depth and diameter can be obtained.
This system can be used together with the automated surface defect removal system developed by ASRC. When a surface defect is identified, the removal system may execute artisan level workmanship to repair the defects. This enables automation of jet engine maintenance procedures and significantly shortens the process from the 4-hour manual method to 1.5 hours, enhancing the quality, reliability and accuracy of work; and cuts down operation costs. It can thus substantially reduce the turnover time for aircraft maintenance and airplane turnaround time.
In future, ASRC will apply deep learning defect detection in other areas such as Non-Destructive Testing (NDT) images or fuselage inspection to further realize automation in aircraft maintenance. In the perceptive of Big Data and Industry 4.0, this kind of automation enables information collection and pattern analysis of defects, leading to better engine part design in future.
Click the link above for the video.
"Pitting" is a surface defect found on critical jet engine components of complex shapes and different sizes. It is currently detected by naked-eye inspection. Such task takes up significant man-hours, and embodies errors in accuracy and reliability which limit engine maintenance performance in terms of cost, time, quality, and safety. The newly-developed automated system by Aviation Services Research Centre (ASRC) of Hong Kong Polytechnic University employs robotics, artificial intelligence and deep learning, image processing, 3D scanning, optical metrology, and non-destructive testing technologies to identify and measure "corrosive pitting" such that precise information such as pitting position, depth and diameter can be obtained.
This system can be used together with the automated surface defect removal system developed by ASRC. When a surface defect is identified, the removal system may execute artisan level workmanship to repair the defects. This enables automation of jet engine maintenance procedures and significantly shortens the process from the 4-hour manual method to 1.5 hours, enhancing the quality, reliability and accuracy of work; and cuts down operation costs. It can thus substantially reduce the turnover time for aircraft maintenance and airplane turnaround time.
In future, ASRC will apply deep learning defect detection in other areas such as Non-Destructive Testing (NDT) images or fuselage inspection to further realize automation in aircraft maintenance. In the perceptive of Big Data and Industry 4.0, this kind of automation enables information collection and pattern analysis of defects, leading to better engine part design in future.
Click the link above for the video.
