Success Story
The Dutch company PackCheck BV offers visual quality control solutions for the food and beverage industries, agribusiness, and logistics. In discussions with customers, PackCheck BV found that they were looking for ways to ensure consistently correct packaging in all their packaging lines. In addition to the packaging quality, this also involves checking barcodes and data codes. In production lines that are not fully automated, quality control is carried out manually and randomly by employees, which results in a high susceptibility to errors. As a result, products with quality deficiencies can end up in retail, leading to product recalls due to packaging defects or incorrect or missing allergen markings on packaging or labels.
To automate the process, PackCheck BV has developed a modular inspection platform for food producers with multiple production lines. The platform uses the ViTrack system control software developed by PackCheck BV, within which the MVTec HALCON machine vision software takes care of image processing. The platform enables companies to inspect many different products in a short time period. The goal is to minimize the number of recalls through inline quality control while gaining better insights into their own processes. Thus, in the event of recurring defects conclusions can be drawn to optimize the production process and reduce rejects.
PackCheck BV customers who use the inspection platform in their packaging process need machine vision for quality control of packaging and for verifying labels and markings. Within milliseconds, the technology can perform a surface inspection, recognize codes and characters and also read them. Users can thus access all product information at any time, such as the packaging quality, location, best-before date or list of allergens. MVTec HALCON was selected as the machine vision software. The decisive factor for this choice was that PackCheck BV was looking for a fully integrable solution when developing its scalable and modular platform. They wanted to be hardware-independent and at the same time have a powerful and programmable software in order to flexibly adapt to different circumstances. In addition to MVTec HALCON with its library of over 2,100 operators, the company also leveraged MVTec's consulting services to address the final structural challenges in image processing for matching applications.
In the modular inspection system, camera systems with up to four cameras and a track controller are used at the relevant points of the production line. The modular design of the platform theoretically allows for an infinite number of camera systems to be used. The system is controlled via ViTrack. To ensure smooth interaction between hardware and software, PackCheck BV relies on the robust OPC UA interface, a machine-to-machine communication protocol for industrial automation that simplifies the networking of devices, automation systems and software applications.
To use the platform, the corresponding inspection program for the product to be inspected is first selected and loaded, and the inspection process begins. As soon as a product enters the camera’s field of view, image acquisition is initiated and the image is stored in the local memory. The tracking controller generates an ID and sends it to the vision controller, where the actual inspection of the acquired image takes place using machine vision with MVTec HALCON, based on the inspection program. Depending on the requirements, machine vision technologies such as blob analysis, matching, barcode and datacode reading or OCR are available. OCR is used, for example, to determine the best-before date and listing of allergens on the label. Barcode and datacode reading determines whether the product in question is labeled correctly. Blob analysis and matching methods are used to inspect the packaging surface by comparing the product against the template. The result is securely transmitted to the tracking controller via the OPC UA interface, among other methods. By matching the result with the ID, it is always clear which product is being inspected. The tracking controller then proceeds correctly with the product based on the inspection result. A defective product is sorted out, while a defect-free one is further processed. In this way, the inspection platform is capable of performing the complex inspection processes at a speed of 100 packaging units per minute.
Since the beginning of 2023, the inspection platform has been implemented by companies in the food industry to automate the inspection of multiple production lines, each producing different products. PackCheck BVs' customers have achieved improved efficiency, better utilization of their workforce, and reduced waste with the platform. The inline quality inspection checks every single product, leading to a significant reduction in product recalls, allowing food producers to respond to the demands of critical consumers. In addition, food manufacturers gain valuable insights into their own production processes, leading to further automation of processes for the benefit of customers.
