Quality Control: detect your product defects on the production line
AI Application COMPUTER VISION SMART FACTORY DEEP LEARNINGUsing Artificial Intelligence techniques applied to Machine Vision, process control is achieved through the accurate, real-time analysis of images and data from advanced sensors, where the traditional approach is impossible. Process quality control systems allow for 100% verification of production, eliminating the subjectivity of analysis and guaranteeing the end customer a product that conforms to requirements.
Some examples?
LABELS AND BARCODE
The system is able to: check the positioning of labels in the packaging and the compliance of any text and information required by law (such as, for example, the lot or expiration date); check the presence/absence of the correct barcode.
PACKING
The system can check the regularity of the packaging (presence/absence of scratches, cracks, dents), conformity of the pre-packaged product and the packaging as a whole by detecting different parameters than those set and automatically identifying: area, length and size of the defect compared to the theoretical parameter
FINAL ASSEMBLY CONFORMITY
The system, through a smart-camera mounted on a cobot, performs visual inspection directly on the line verifying that the assembled product contains all the necessary parts for product compliance. The solution, together with the presence of the human operator, leads to accuracy close to 100%.
SURFACE SHAPE AND FILLING
The solution verifies: the macro surface aspects that impact the final quality of the product (such as presence and/or absence of burn marks); the shape of the product and that the perimeter of the product does not protrude from the package prior to packaging; the filling consistent with “preset recipes”.
CHECK FILLING
The solution, through a camera at the edge of the line, verifies that the filling of the contents inside a package conforms to the preset quantities. If it detects a shortage or surplus, it reports the error so that the operator can remove the failed piece from the line.
DEFECT COUNTING AND REPORTING
Verification of product characteristics and any product anomalies that compromise product quality and end-customer satisfaction. Non-contact real-time quality control is followed by automation to eject unsuitable products without requiring any human intervention, reaching an accuracy of 99.99%.
beanTech & NVIDIA, of which we are Elite & Metropolis partner, are your allies for quality control with Artificial Intelligence.
