Case Study: Niko switches achieves smarter quality control and fewer false rejections with Robovision and delaware BeLux

Niko switches to AI-driven quality control with Robovision

Niko switches, a leader in home automation and switchgear, wanted to improve the quality control of faceplates and other finishing materials at its Sint-Niklaas facility. Its partly manual inspection process was causing too many good products to be rejected and some defective ones to slip through, increasing costs and reducing efficiency. delaware BeLux worked with Robovision to support a more accurate, AI-driven inspection approach.

delaware BeLux implemented Robovision’s computer vision and deep learning platform on Microsoft Azure to automatically classify defects from camera images, training it with a few hundred examples of perfect and faulty faceplates. The solution detected defects at pixel level and did not require manual programming, helping Niko reduce false rejections of good products by 72% and cut the number of bad products that remained on the conveyor belt by 33%.

Niko switches

Jonathan Berte

Chief Executive Officer