The challenge
What needed to change.
The filling system relied on time-consuming manual checks and handled varied jar formats containing products with different visual characteristics.
Vision & robotics · Food, beverage & packaging
Automated jar inspection
Automated inspection for different jar sizes and products, reducing reliance on manual quality checks.
ClientInternational food manufacturerSectorFood, beverage & packagingPrimary disciplineVision & robotics
Adbro’s approach
How the solution was engineered.
Adbro combined repeatable lighting, product recipes and automated checks for key packaging and fill attributes.
The outcome
What the project delivered.
Quality decisions are faster and more consistent, with less manual effort and broader process coverage.
Technical delivery
Inside the engineered solution.
The detail below shows how the main equipment, software and interfaces were brought together—not just what the project achieved.
150jars inspected per minute5colour cameras per inspection system360°view of each jar2systems installed
System scope
- Five Point Grey Blackfly colour cameras
- Polarised white-light illumination
- Height-adjustable top camera
- IFM product sensor and SICK conveyor encoder
- C# application using HALCON image-processing tools
- Recipe-based HMI configuration
- Pneumatic reject mechanism
- Statistical database, image export and spreadsheet reporting
Multi-view image acquisition
Four side cameras capture the jar around the x/y plane while a fifth, vertically adjustable camera views the top. Polarising filters control reflections from the glass, labels and white LED lighting.
Position-linked capture
An opto-reflective sensor detects each jar and a conveyor encoder tracks its movement at up to 0.5 metres per second, linking image capture and rejection to the correct product.
Configurable inspection software
Bespoke C# components coordinate acquisition, product location and reject decisions. Product-specific image-processing classes use HALCON routines, with settings exposed through the HMI under recipe control.
Quality information
The system records reject categories and inspection statistics, can export production images for remote analysis, and produces spreadsheet data for production and quality teams.
Verification & handover
Commissioning covered each jar and product format, checking camera position, recipe selection, feature detection, encoder tracking, timed pneumatic rejection, data logging and recovery after stops or product gaps.
Project highlights
Pack inspectionRecipe handlingMultiple formatsFood quality
Start a conversation
Could we engineer a similar outcome for you?
Bring us the challenge. We’ll help you shape a practical next step.