Step by step, we are developing an AI model to detect damage on metal surfaces. For this project, Syrine Zbdidi is using machine learning to scan metallic tool blocks for damage. These tool blocks are used for composite part production in the aerospace manufacturing industry.

Syrine: “Our assignment consists of several aspects. First of all, we are training the robot to autonomously detect where the tool blocks are located on a cassette. Then, the robot needs to accurately scan each tool block, starting and ending at the right position and maintaining a steady distance for high quality inspection. Subsequently, the system assesses the images. Are there any scratches, dents or other anomalies? The outcome must be of a high reliability in the end.”

Watch the video below for the progress!