Needle and Syringe Assembly Verification
Ensure an adequate connection between needle and syringe
Prefilled syringes have the needle pre-attached, or staked in. The end of the syringe cylinder is heated and a cone, or nozzle, is formed. The cannula of the needle is then inserted into the cone and secured typically with a UV-cured adhesive. For a polymer syringe, the needle is affixed via insert molding.
The connection point of the glass or plastic of the syringe with the needle needs to be inspected to ensure that there are no chips, air bubbles, or other defects, and that the amount and placement of the adhesive is correct. The appearance can vary depending on adhesive type, and this can change over time and between lots.
In the past, difficulties with needle bonding adhesive inspection have limited the range of gauge and length possibilities for staked needles.
Needle and syringe assembly inspection is best performed with a combination of Cognex Deep Learning and traditional vision algorithms. The various dimensions of the cone, such as diameter, length, various angles, and planarity of the end, can be easily measured with a standard machine vision system, such as the In-Sight 8505P.
However, inspecting components through reflective transparent material, whether glass or polymer, is a task for deep learning technology. Cognex Deep Learning is trained on a range of acceptable nozzles with inserted needles. Any needle and syringe assembly that exhibits properties outside of the acceptable range, which could include bubbles, cracks, inadequate needle bonding adhesive, problems with the cone, or other inclusions, is flagged as a defect and rejected from the line. Since Cognex Deep Learning can be easily trained on new needle lengths or gauges, manufacturers avoid lengthy and complicated re-programming that would be experienced with conventional machine vision.