Loading…
Back To Schedule
Wednesday, June 2 • 14:50 - 15:15
AI driven quality assurance for composite parts using ultrasound

Sign up or log in to save this to your schedule, view media, leave feedback and see who's attending!

  • Business case behind AI
  • Deep Learning assisted defect detection
  • Wider applicability

Premium AEROTEC has developed a deep learning model for defect detection in composite aircraft components based on ultrasound scans. Non-destructive testing is indispensable for the production of CFRP parts in the aerospace industry. For large components, such as the A350XWB fuselage shells, ultrasonic testing is the standard method of detection. Acquisition of data for these parts is already highly automated. Data evaluation is nevertheless fully manual and may require more than one working day per part. Continuous tasks demanding human expertise, to be applied repeatedly across the complete scanned area of each part, requires AI assistance to increase output whilst maintaining quality check standards. The business case for reducing manual overheads, through the automation of data evaluation, mitigates fatigue and other phenomena that decrease quality.
Classical computer vision methods are insufficient to accomplish this goal. With the recent success of artificial intelligence in the field of image recognition, it is now possible to automate defect detection. A general approach to automation using Deep Learning will be presented. A wider applicability of the method to other signal types such as X-ray, CT and active thermography is possible.

Speakers, Jury Members & Final...
avatar for Olaf BEESDO

Olaf BEESDO

Data Scientist, Premium AEROTEC GmbH
Olaf Beesdo has worked for many years within the aerospace industry with particular interest in industrializing the latest technological trends. Currently, he works at Premium AEROTEC GmbH as a Data Scientist. His remit is the analysis and optimization of manufacturing processes with... Read More →


Wednesday June 2, 2021 14:50 - 15:15 CEST
Conferences (Hall 6 - Room 611)