AI-based error detection at Emmi
Emmi utilises state-of-the-art technologies, including artificial intelligence, in its production facilities. On behalf of Emmi, soxes developed a prototype for error detection in the yoghurt production process.
Efficient quality assurance
The prototype checks the yoghurt pots at the end of the production chain. A camera photographs the lids and an app checks whether the correct lid has been used.
Technologies used
The solution is based on Microsoft Custom Vision, an Azure service for image recognition. A Python app processes the images and compares them with product data. In the event of errors, the camera sends a warning signal.
Fast, agile implementation
Thanks to intensive collaboration and agile methods, soxes created an efficient solution for Emmi within 6 weeks, which brought great added value.