Babynov introduces AI-assisted vision into its quality inspection of cans

Did you have any apprehensions about artificial intelligence?

We had no automated inspection equipment combining vision and AI on our lines; this technology was unknown to us. So we took the time to clearly define the goals and the expected results. We also had many discussions and worked closely with Psycle to make sure we reached the expected level of performance. In the end, the results are there: we are satisfied and we have decided to automate other inspection stations with Psycle.

We measured a significant reduction
in our downtime

What were the key points of your first project?

We worked on two points that I believe are the keys to a successful project: AI-based recognition of defects on cans, and the mechanical integration of a system to eject the products identified by the AI as defective. Psycle brought its expertise on both points, which allowed us to be particularly efficient.

Psycle's AI gives us immediate responsiveness in detecting defects.
The interface is easy to use and allows instant data analysis.

les solutions psycle

Démarrez votre projet de vision industrielle avec le SDK Psycle

un framework Python compatible avec les standards de vision (GenICam) et les derniers modèles de Deep Learning.