Identification and Classification of Defects in Aluminum Strips

Project Metal Inspection Engineering, 2022–23

Liaison(s): Jeff Foutch, Andrew Donelick ’15
Advisor(s): Nancy Lape
Students(s): Nick Casañas (TL-S), Sidney Taylor (TL-F), Olivia Tuffli, Viviane Solomon, Madeleine Masser-Frye, Brandon Bonifacio, Carlos Sanchez

One of the leading aluminum manufacturers in the nation has requested that the Clinic team develop a system to identify any potential defects on the surface of aluminum coils. The team designed an image-collection system to collect images of the product as it passes through the manufacturing line. With those images, the goal of the project is to develop a machine learning computer vision system that can detect and flag potential defects in real-time during the manufacturing process.