Chamzinskaya Poultry Farm (Nord Clan: ML Sense)

2024: Implementation of ML Sense system in chicken meat production

Chamzinskaya Poultry Farm JSC (Good Business Agro-Group) is among the top 10 poultry producers in the country. Last year alone, the company shipped 156 thousand tons of broiler meat. The poultry is sorted after treatment after cooling.

Before the automation of the quality control system using machine vision from ML Sense, poultry specialists manually rejected chicken with defects.

ML Sense is a digital platform based on machine vision and neural networks for product quality control at conveyor-type production facilities (included in the register of Domestic software). The project implementation period at the Chamzinskaya Poultry Farm was only 3 months.

Stages of system implementation in the enterprise

1. Design and installation of software and hardware complex in production.

We selected video cameras, lighting fixtures for more accurate recognition of defects. They were installed on the fastening systems in the cooling workshop. They provided two cameras on either side of the conveyor: one camera looks at the breast of the chicken, the other back.

𝓲

Nord Clan

2. Collection of data set and marking of defects in photographs.

We took photographs and took 5,000 photos of bird carcasses hung on the conveyor. Then they marked all the defects on them: hematomas, cuts, broken wings. They trained the neural network to evaluate objects and respond to defects. At the same time, it was necessary to take into account the dimensions of the defects: cuts - from 10 mm, hematomas - from 300 mm.

𝓲

Nord Clan

3. Controllers were set up and installed, which send a signal to the culler in the workshop.

The system works like this: when machine vision sees a defect, a signal is sent to the controller. In turn, the controller sends a signal to the culler, which removes the defective carcass from the conveyor.

4. Launched the system into operation, trained personnel.

While working on the project, a team of Nord Clan engineers went to production several times to test the operation of the system. And only after both parties were convinced that the system works stably and without failures, we handed over all the equipment to the customer for operation, trained personnel, signed acts of acceptance - transfer.

𝓲

Nord Clan

Result

The case of Chamzinskaya Poultry Farm JSC is another example of how you can automate the product quality control system using machine vision. Detection of defects with ML Sense is carried out with an accuracy of 99%. The production of defective products is reduced to zero.

What has changed in the plant:

• Visual inspection is replaced by machine vision technology. Quality control work is faster and more efficient.
• Increased the economic effect. Specialists no longer need to manually reject the non-condition. This means that they can do other work, which ultimately reduces the wage fund.
• They introduced domestic software, which means they solved the issue of import substitution at an industrial enterprise.