USB Connector Inspection
Ensure that USB-C connections are functional
Graphical programming environment for deep learning-based industrial image analysis
USB Type C connectors are replacing various existing USB types as well as other connectors for power and data transfer. They are small and need to be manufactured to tight specifications. USB-C connectors can handle much higher power than previous USB models, but are also smaller, with thinner walls, making them easier to damage during manufacturing and raising the risk of electrical problems and even fire if there are defects.
Before the connector is attached to the cable by a molded holder, all sides of it need to be examined for defects, for a total of seven sides, including the interior. Possible defects include scratches, dents, deformation in a variety of locations, dust, scratches inside, and excess glue. Many of these defects are difficult to spot at the speeds and volumes necessary for inspection. Even small defects on the connector can affect the quality of the electrical connection, how tightly the connector is held in the socket, and how easy it is to remove again. These USB connectors must last for the life of the phone.
The current defect detection procedure involves human inspectors using microscopes or traditional rules-based machine vision-based AOI machines where operators view image captures on a monitor.
A combination of Cognex Deep Learning’s defect detection and classification tools can dramatically improve USB-C connector inspection. The defect detection tool trains on a set of labeled images that provide examples of both good and bad connectors. After this, it detects and flags any anomalies while passing connectors with minor cosmetic marks that do not affect function.
The classification tool trains on a variety of functionally significant variations in appearance and on visual variations that do not have a functional effect. It improves detection accuracy, while also providing information on recurring errors that helps identify potential performance problems in machinery upstream and supports continuous process improvement.