Our AI technology performs reliable automated visual inspection that helps to detect scrap early in the production process while providing real-time insights thanks to quality tools for traceability, analytics, and monitoring. The solution proved itself even in environments where traditional deep learning solutions fail – e.g. in presence of liquids, dirt particles, and reflections.
The system inspects the whole surface of the ring, including the inner diameter, the outer diameter, and the faces and detects a wide variety of surface defects (down to 0.25 mm in size), such as:
The technology also provides the richest data about every single produced part. These are archived, analyzed, and visualized to truly enable data-driven total quality management:
We go far beyond just training a deep learning model on an annotated dataset. Coping with real-world uncontrolled environments, setting up and maintaining the learning loops, artificial data generation, software aided annotation, scoping the MVP…you name it.
Since 2016 we have built 30+ AI & machine vision solutions and advised on hundreds of projects. Combining the best of research and our technologies with best practices in business-ready solutions development, we give your AI transformation a direction.