Skip to Content

ONLINE 3D INSPECTION OF COOKIES: A SOLUTION FOR THE FOOD INDUSTRY

In the food industry, finished products must meet strict visual quality requirements, especially on high-speed production lines. Cookies, produced in large quantities, must have a uniform appearance to meet marketing criteria and satisfy consumer expectations. A visual defect, even a minor one, can lead to rejection during final inspection, material losses, or impact the brand image with the end customer and distributors.

THE CHALLENGE

The challenge of this project was based on the need to perform a complete integrity check of the cookies, including:

  • Measurement of diameter and ovalization
  • Height assessment
  • Identification of the type of cookie (face or full)
  • Detection of material variations (deficiency, excess, clumping) Control of the orientation of the biscuit

All of this had to be accomplished with a very high production rate. Specifically, 50 cookies had to be inspected every 600 milliseconds, with 6 checkpoints per cookie: a total of 300 checks performed in less than one second.

THE METHODOLOGY

Hardware architecture

  • 2 × LJ-X8900 placed side by side to cover 1,500 mm on the X-axis (bandwidth) with ~50 mm of central overlap.
  • Blue laser 405 nm, 3200 points/profile (mode "Parallel Imaging").
  • XG-X2900 Controller (Multi-core CPU + Dedicated 3D FPGA)
  • Incremental encoder
  • Complete Machine (Mech + Elec + Automation) made by SG2C
Process parameters
  • Y Axis (scrolling): 30 m/min = 500 mm/s.
  • Acquisition time + processing: 600 ms per image.
  • 50 cookies inspected per image
Resolution & measurement window
  • Resolution X (bandwidth): 0.225 mm/px.
  • Y Resolution (scrolling): 0.225 mm/px.
  • Reliable detection of a chip ≥ 1mm² and a chocolate thickness variation of ± 1 mm.
Measured performances
  • 50 biscuits × 5 algorithms × 100 cycles/min ≈ 25,000 tools/min.
  • CPU Load XG-X2900: ~45% at 30 m/min → margin for 45 m/min.
  • No daily recalibration; integrated Z drift self-diagnostic.
Treatment chain
  • Decision & action :
    • Unit ejection via 16 individual nozzles
    • Complete ejection if too many NG cookies due to conveyor retraction

THE RESULTS

Thanks to this solution, the online quality control of cookies achieves a guaranteed compliance of 99.8%, while optimizing production:

  • Reduction of false rejections, improving efficiency and limiting material losses.
  • Maintaining the OEE (Overall Equipment Effectiveness) through unit ejection.
  • Real-time tracking, with a precise traceability system and integrated automatic diagnostics.
  • Optimal performance, with controlled CPU load allowing for an increase in pace if necessary.

The integration of this 3D vision technology allows the food industry to ensure impeccable quality while maintaining a high production throughput. With customized solutions tailored to industrial constraints, CODA Systems supports companies in optimizing their quality control processes.

ANY QUESTIONS ?

If you would like to share a project or need more information, our team is at your disposal.

 CONTACT ​​CODA​​​​