Enhanced Clarity for Smart Vision

Food surfaces vary widely, which can overwhelm many smart camera systems that typically struggle to distinguish between quantitative and qualitative features. As a result, product defects may go unnoticed, or defect-free products might be misclassified as faulty. This leads to inaccurate quality control, high costs, and operational down times. That’s why Sensaray asked us to develop a Minimum Viable Product (MVP) to significantly reduce the error rates of their inspection systems.

We made the involved systems smarter — using artificial intelligence. The cameras can now better determine whether a product meets quality standards, regardless of its individual surface. We trained neural networks on high-performance computers to make precise quality decisions directly on the production line. These decisions are then communicated to the machines via various protocols. The result is the ability to quickly and accurately assess production quality and sort out defective products.