Vision Guided Robotics for Laser/MIG Welding Seam Tracking

Weight reduction has been one of major driving forces of research and innovation in automotive industries. Changing from the conventional Laser overlap welding to edge welding could save 50% of total weight of joint flanges of workpiece. Also, edge welding can significantly improve the welded joint mechanical properties. However, edge welding is challenging since the thin laser beam needs to be guided on the joint constantly within tight boundaries. For this reason, project of vision-guided robotics for laser/MIG welding seam tracking is proposed. The idea is to use a vision sensor to track the joint constantly and accordingly correct the laser welding TCP trajectory when any deviation is detected. The research mainly includes finding a suitable vision sensor, designing an algorithm of noise filtering and edge detection, examining approach of laser TCP trajectory correction, and implementation of system integration. Benefits of a vision-guided robotic laser edge welding system will be reduced weights and higher quality welds.

Faculty Supervisor:

Fengfeng Xi

Student:

Yu Lin

Partner:

Van-Rob Inc.

Discipline:

Engineering - mechanical

Sector:

Aerospace and defense

University:

Ryerson University