Brisbane meatworks finds ideal vision solution

When your business supplies fresh meat products to major supermarkets and consumers across Australia, you can’t afford to get it wrong. Products displaying illegible date codes, damaged packaging or wrongly labelled products can be disastrous for suppliers.

The Brisbane meatworks is unique. It is a globally recognised fully integrated facility that completes a full circle in beef production including slaughter, boning, value-add, retail-ready and distribution. According to its production manager, the site processes some 1200 head of cattle per day.

Apart from bulk meats it also produces stir-fry and diced beef and veal, beef sausages, corned, marinated, glazed and coated beef and veal products, corned beef silverside and hamburger patties.

“We currently have an annual production of over 15 million kilos for national distribution to prominent retail and supermarket shelves,” said the production manager. “Due to very stringent requirements demanded by our retail distribution partners, our entire packaged shelf-ready product needs to be exactly as per what is ordered and labelled as.”

The biggest challenge for any meatworks is traceability. Most facilities traditionally rely on casual labour in the final packaging and inspection process.

An inspection system was therefore needed that could guarantee speed, product integrity and accuracy. The meat works commissioned Omron, through its integration partner Pac Technologies, to install a vision inspection system that would minimise the risk of sending out any non-compliant products.

A challenge for this application was to check both 1D and 2D barcodes at varying focal lengths on the final production line. Also, there was a requirement to check and verify the date code on each shelf-ready product.

Kim Simonsen from Pac Technologies, in conjunction with Omron application engineers, created a vision solution using FQ2 machine vision cameras on each line. On some lines, two cameras were used at varying focal lengths to handle the varying heights of the target product.

The carton barcode is pre-checked to verify that the product is as expected before the individual packs are checked.

“We accessed some very powerful algorithms built in to the FQ2 camera to achieve what the customer needed to satisfy their date and barcode checks,” said Yang Qui, a senior application engineer form Omron Electronics, Brisbane. “OCR was used to not only check for the presence of the date code, but actually read the text to ensure that the date code was correct and readable. The small-sized 2D barcode was a challenge and required us to employ the high-resolution version of the FQ2 vision camera to obtain a reliable and accurate reading each and every time.”

Omron Queensland State Manager Paul Gibb said Omron is also assisting other Queensland-based meatworks that produce down to shelf-ready product as well as bulk packs.

“There are common issues emerging when talking to each company about checking integrity and accuracy of the final packaged product and its labelling and identification,” he said. “Hundreds of different product and label variants, many types of barcodes and date codes, varying existing PLC architecture, high turnover of transient workforce and a hostile operating environment all present a challenge to a solid and reliable vision solution.”

One of the main challenges was how to process the data once reliable and accurate judgements of the final product are obtained.

“In this instance we used our powerful and flexible NJ Machine controller with SQL connectivity and EtherNet/IP to communicate directly to the customer’s database without the need for any software-based middleware,” said Gibb. “Since Omron’s NJ controller has the option of EtherNet/IP communications, it communicated directly with the customer’s existing PLCs, creating a seamless network from camera to database.”

Gibb said Omron’s FQ2 vision cameras are rugged enough to be installed directly on the production line in a meatworks hose-down environment and they have enough capacity to store more than the customer’s total product line-up and label varieties.

They support up to nine types of barcodes. The Australian meat industry also widely uses GS1-databar code and the FQ2 cameras have been successfully used for this type of product verification and production information inspection.

Originally published here.