Liquid Injected Molded (LIM) Rubber Seal Inspection
Detect small defects in flexible waterproof seals used in smartphones
Graphical programming environment for deep learning-based industrial image analysis
Flexible, molded seals of liquid injection molded (LIM) silicone rubber are used in many areas of a smartphone to protect sensitive electronic components from water. Examples include perimeter gaskets running around the perimeter of the phone to seal the front and back panels, screen seals, and rubber seals used under the speaker mesh and around charging ports. Since a defect anywhere on the seal will compromise waterproofing, any foreign inclusions, press damage, excess glue, or other defect must be detected before final assembly.
Inspection involves the use of multiple cameras to provide a full 360-degree view of the seal’s sides. From a side-on view, the seal is a narrow strip on a complex, confusing background. Traditional machine vision can’t detect all the possible defects while passing the natural variations in the flexible material.
Cognex Deep Learning defect detection trains on a set of images that identify the full range of good silicone rubber seals and on a set of images highlighting defective seals. Even when compressed, the flexible seal can vary in appearance without compromising waterproofing. The defect detection tool has no trouble with the variation in good parts and can pass those while identifying anomalies that do have a functional consequence.