Product Recognition in Pharmaceutical Logistics

Ready Made

REACH-2021-READYMADE-COFARES_2

COFARES

To check whether an order is fully and correctly picked using images taken during the order picking process. It should detect whether there are differences between the expected products and the real ones.
There are some of our logistics processes, such as the detection of differences between the delivery note and the delivered products, that are automatic and involve recordings or photographs, but they usually need human supervision. The goal of this challenge is to develop an artificial vision tool accurate enough to dispense with human supervision.

The orders placed from the pharmacies are picked into plastic totes. They are moved onto conveyors from picking station to picking station. At each picking station an operator or an automatic system puts the products into the tote and an image is taken. Finally some of the totes are checked at a checking station and an image is taken by each product.
Currently those images are used in case a customer claims for missing products. An operator looks for the missing product in all the images and if it is found the customer is informed that the product was delivered.
In the future those images could be used in a proactive way to detect missing pieces before finishing the picking operation.

To automate the process of looking for missing products so that the operator stops doing and only supervises and solves specific problems.

Any improvement suggestions about the physical environment and the devices used (cameras, lightning, etc.) will be welcome.