Medical Device Part Inspection
Reduce recalls and rework costs with machine vision and deep learning
Medical devices come in many complex shapes, sizes, and surfaces, from a shiny metal knee replacement to the small webbing of a stent. The use of additive manufacturing makes geometric shapes even more complex for certain procedures. Because medical devices are used on or inside the human body, quality inspection of the part is critical. Pacemakers, catheters, scalpels and other medical device and surgical equipment could have microscopic surface defects, scratches, burrs, dents, or contamination that could be harmful to a patient.
Many medical device manufacturers rely on human operators or rule-based machine vision systems to ensure parts and plastic components conform to quality and safety standards. Relying on human quality inspection can be costly and sometimes allows defects to slip by. Machine vision systems are great with a known set of variables in a controlled environment, but the variability of medical device shapes and surfaces makes some inspections impossible to solve with machine vision alone.
In some applications, like measurement, rule-based machine vision will still be the preferred and cost-effective choice. For complex inspections involving wide deviation and unpredictable defects, deep learning-based tools offer an excellent alternative. Cognex deep learning solutions locate, analyze, and enable classification of complex inspection issues to stop defective products from entering the supply chain. Deep learning combines the human-like inspection capabilities with the automation and repeatability of a computerized system. This can be augmented by the use of robotics to ensure machine handling and vision tools work together to inspect the most complex anomalies sometimes missed by operators. The end result is less recall events, lower rework costs, and full image capture and traceability.