Nutrunner Tracking in Production Line with Image Processing

At a Glance:

  • Real time monitoring and detection quality faults

  • Comparing with CAD data for dimension check

  • Painted/Unpainted product quality check

  • Plastic and Metallic product quality check

  • Transparent/Glass product quality check

  • Product cut view quality check

  • Implementable into though production workplaces

  • Low hardware cost and friendly UI



Solution is directly focused to check the quality faults during the production. With Deep Learning and Machine Learning algorythms, system has been checked and detected the quality faults on the product surface due to painting or mould marks or poor workmanship.


The system should meet the following requirements:


  • Detection of quality faults on the painted plastic products

    • Less paint

    • Paint bump

    • Micro scratchs

    • Paint curtaining

    • Orange peeling on the paint etc.

  • Detection of quality faults on the unpainted plastic products

    • Micro scratchs

    • Mould marks

    • Shape control

  • Detection of quality faults on unpainted metallic products/sheets

    • Micro scracths

    • Mould marks

    • Notch control

    • Flash control

  • Detection of quality faults on transparent or glass products 

  • Product dimension measurement

  • Dimension measurement compaing with CAD data

  • Product recognition in the production phases


Teach the System with your Quality Metrics


System uses deep learning and machine learning algorythms to detect the product quality faults in the production. System provides an early warning mechanism to prevent the increasing quality faults before it is soo late by this means reject product costs will be reduced. 


System also needs to understand and learn the quality standards/thresholds of the customer with deep and machine learning user friendly tools. After complition of this phases, system will be answering and guiding the quality teams fully automatic decisions. 


Image processing cameras can implement with different aspects to catch the best view for %100 quality detection ratio.