AIVisualMRO is an advanced AI platform designed to help inspectors complete visual inspections more efficiently and consistently. It combines audio transcription, computer vision, and large language models to support every stage of the inspection workflow while ensuring that the human inspector always maintains full control over the final decision.
Surprisingly there are lot of compressor and turbine blades for sale on e-bay. I managed to buy some compressor blades from a McDonnell Douglas AV-8B Harrier II, in a very good condition.
Managed to build & train a YOLO 11 on 3 different types of compressor blades. The dataset was
build from scratch and augmented in order to improve teh detection accuracy, while keeping the enviromental
conditions limited. The today time for training was 1 day including building the dataset. This serves as a
demo/usecase that this kind of training can be done very fast, and can give results within hours.
It helps itterate fast and figure out if any problems will appear later in production.
On the main page showing a video which is a small segment of the video inspection and detection of parts.
Below there is another video showing blade segmentation of an industrial turbine engine.
The AI model has been trained to 'extract' the blade part of the turbine part and ignore the root.
Once the blade image has been extracted the coating defects can be detected, i.e. missing coatings.
Each image/blade is saved in the final report, toghether with the % of coating left, and the technician audio comments in text format.
Each image/blade is ID-ied and stored in a database for later analytics.
