Apetito has been providing frozen foods and ready-to-eat meals to schools, nurseries, enterprises, hospitals, and retirement homes throughout Europe and Canada since 1958. The company’s meal trays are made by placing various food products onto trays for shipping and delivery. Customers’ complaints about missing food items from trays prompted the corporation to look for an automated way to inspect finished products.
In an attempt to address the issue, Apetito began weighing each tray as it exited the production line. However, this did not fully fix the issue because if one of the food items on the tray was somewhat heavier than usual, it may be mistaken for what was missing. Apetito also intended to improve other manufacturing processes, such as making sure each lid on its desert containers was properly crimped and secured.
One of the first initiatives Apetito made to automate visual inspections was to monitor one manufacturing lane with a Raspberry Pi-based system. Models attached to an output signal in the control system that would push rejects off the line were informed by photos of good and undesirable items. The company saved more than £15,000 per year in labour costs by not needing a worker to conduct this job.
The issue with the Raspberry Pi system is that it required hundreds of images of its items to be uploaded to a USB and then transferred to a computer over the course of three to four hours to train the system. The information gained has to be relayed back to the line. New training methods for the Raspberry Pi system were necessary whenever Apetito made adjustments to its meal production operations to maintain adequate stock levels.
Be First to Comment