Automated real-time inspection for robotic arc welding operations

Welding has widespread use in manufacturing, and its quality often determines the overall performance of the manufactured part. Nevertheless, due to the complex nature of the process, weld quality inspection remains manual with the standard approach of QA for defects after production completes. This introduces significant costs of downtime and reworks associated with finding defects at the late stages of production. This project leverages advanced data analytics and machine learning to develop and validate an automated real-time quality inspection system for industrial welding operations. Manufacturers greatly benefit from real-time insight about the welding operations enabling them to avoid defects early during production.

Intern: 
Seyyed Mohammad Mike Mohseni
Faculty Supervisor: 
Daan Maijer
Province: 
British Columbia
Partner: 
Sector: 
Partner University: 
Discipline: