Platform for the food and beverage industry
Automatic inspection with machine vision

Automatic inspection with machine vision

Securing quality and increasing efficiency

In the food sector, both efficiency and quality assurance are essential. An error in the production chain - for example due to a different shape, size or texture, a wrong label or physical contamination - can lead to rejected end products, recalls, image damage or health risks. Automatic quality inspection with machine vision, AI and innovative technologies secures quality, minimizes errors and ensures efficient production.

With machine vision systems, it is possible to collect detailed data about the quality of each product on the production line. This provides valuable information for quality improvement and compliance. "Automatic inspection is done at high speed. Typical image capture times range from 0.1 to 0.2 seconds per camera. This means that vision systems can inspect up to hundreds of products per minute," says Paul Bakker, account manager at EKB, which specializes in total industrial automation solutions.

3 copy
Conveyor at the plant for the production and bottling of juices in cardboard packaging.

Vision systems are fast and accurate. "The human eye does not see everything. Thanks to early, fast and accurate identification of deviations, defects or contaminations, you can reduce waste and returns. This leads to cost savings and consistent high quality. It contributes to customer loyalty and reinforces brand trust." 

Transparent film

Tycho Scholtens, business developer vision at EKB adds: "For the human eye, assessing food such as cheeses or meat products packed in transparent foil is sometimes difficult. For this we have developed a technique so that the foil is not an obstacle to checking the quality of the food. It also allows our customers to determine defects such as spots, shape, texture and color defects of already packed products."

Artificial intelligence

Artificial intelligence (AI) also contributes to automatic quality inspection. Scholtens: "If you make products according to a fixed recipe, it is better to program this using rule-based vision. However, if you are dealing with small volumes with higher variability, then you can use an AI model to teach the system to deal with these variations. This is easier than reprogramming."

1 copy 1
The vision system at DOC Cheese automatically checks the top and bottom of the cheese for contamination and damage to ensure consistent product quality before the cheeses leave the factory.

Flexibility versus security

Scholtens explains this with an example: "With chocolate bars, there are various deviations in the structure that are not easy to describe in a few programming lines. With an AI model, you can properly assess deviations in shape, structure and color with much less effort so that the product quality is consistent. It is flexible and easy to implement. However, proper training is necessary. The reject rate is often higher than in applications set up in the traditional way where you can set the reject limit. So both options are possible. Using programming language leads to more certainty, while with AI you create more flexibility. Often a combination of the two is the best solution." 

Thorough preliminary study

In other words, AI is used according to the application. The customer's goal is always paramount. Bakker: "To find out whether a solution is possible, we start with a quick scan. If possible, we will take a number of the customer's products to our vision lab to perform a number of tests. This way, we get an indication of the complexity of the issue and can give our customer a general idea of the solution, the investment involved and what value it will bring to the customer. If the solution and investment fit the customer's business case, a proof of concept follows. This also tells you immediately whether it is technically feasible to achieve the desired accuracies or speeds and to integrate the system into the production process and link it to the machine control or MES system. You thereby take a lot of risk away from the customer before the project starts."

4 copy
Raw meat production

EKB also remains involved during and after delivery. Bakker: "We take care of the process from A to Z and integrate machine vision systems seamlessly into the production process. In addition, our 24/7 service desk is ready to immediately solve any issues that may arise. In other words, we don't just offer products, but services and total solutions that unburden customers."

Heeft u vragen over dit artikel, project of product?

Neem dan rechtstreeks contact op met EKB Group.

EKB logo Contact opnemen

Stel je vraag over dit artikel, project of product?

"*" indicates required fields

This field is for validation purposes and should be left unchanged.
EKB logo Telefoonnummer 0251 - 261 920 E-mailadres [email protected] Website ekb.co.uk

"*" indicates required fields

Send us a message

This field is for validation purposes and should be left unchanged.

Wij gebruiken cookies. Daarmee analyseren we het gebruik van de website en verbeteren we het gebruiksgemak.

Details

Kunnen we je helpen met zoeken?

Bekijk alle resultaten