Integrated product inspection system

An integrated inspection system is designed to detect defects and determine geometry of various sides of small-sized products in continuous production when performing the following tasks:

  • detect the required product defects (chips, scratches, dents, foreign particles, discolorations, etc.);
  • determine the product geometry (linear and angular dimensions with an accuracy of 10 µm);
  • automatically sort the products according to the fail/pass criterion or by color, size and other required criteria;
  • automatically count the sorted products and cut-off by amount/weight using actuators;
  • collect and store the acquired data statistics.

Appearance of the integrated inspection system versions:

Typical system configuration:

  • feeding module (motor-driven feed wheel feeding the product from the feeder to rotary table);
  • positioning module (motor-driven glass rotary table for product positioning in the camera’s field of view);
  • photo module (high-aperture lens camera and proximity sensor);
  • lighting module (dome-type LED light source);
  • sorting module (air-operated valve with a nozzle and two containers for defective and good products, defective products are blown by a compressed air jet into the scrap container and good products are transferred into the good product container through a guide bar);
  • control unit module (electrical control cabinet where the following components are installed: power supply, terminal strips, digital input/output control module, computational server);
  • control panel module (operator touch panel with at least 12 inch responsive screen);
  • frame module (the frame is made of interconnected structural aluminum shapes).

Modular design allows further expansion/upgrade of the following modules:

  • change from a single camera (one side of product inspection) to multiple camera system (from 2 to 8 cameras) in the photo module; availability of a transparent glass rotary table allows to take photographs of the product base parts;
  • install a module to count good and defective products not only by software-based, but also by hardware-based method using an air-operated limiter (e.g. for automatic product picking into boxes);
  • use defect type sorting module (based on deep learning algorithms); the additional containers shall be added to the sorting module.

The final requirements for appearance, design and configuration of system modules are specified in the Terms of Reference.