Deep Learning in Weld Inspection Registration

Duration:30 minutes

Abstract:

When it comes to spot welding, electrical testing will catch some failure conditions, but is insufficient by itself. Visual inspection is challenging and often not done because of the unpredictable variations in both good part and bad part appearance.

Deep learning addresses two key issues. This information can be dynamically fed to a process control system that can trigger a corrective action process

During this webinar, we'll dive into deep learning and how it can help:

Differentiate between good and bad parts allowing for 100% automated visual inspection dramatically reducing production of defective components
Classify the type of defect of bad parts which can then be fed into a process control system triggering corrective action processes
Achieve a level of consistency and cost effectiveness that's not possible to achieve through manually inspection

Speaker

John Petry

Director of Marketing for Vision Software

Cognex

John Petry is the Director of Marketing for Cognex’s Deep Learning Software business. In this role, he works with customers to understand how to successfully deploy Deep Learning machine vision in a factory environment. Prior to that, his focus was on developing Cognex’s consumer electronics business in the Asia-Pacific region.

John has worked in the machine vision business for over 25 years in a range of engineering and marketing positions. He’s been based out of Cognex’s headquarters in Natick, Massachusetts, and the Cognex office in Tokyo, Japan. He holds a Bachelor of Science degree from M.I.T.