Quality assurance: IT-Logix launches Machine Vision offering

"Based on orders from commercial and industrial companies, we have developed a structured approach including a technical architecture plan for the integration of typical peripheral systems," says Samuel Rentsch, Co-CEO of IT-Logix. "This is because we are sensing an increasing demand for machine vision with advanced machine learning methods, in order to also be able to handle complex decision-making processes in the Quality assurance to cover."

Many existing systems have reached the end of their life cycle, so that solutions for demanding applications are being evaluated for new procurement. The range is complemented by a qualified selection of camera products and cloud services from various providers to generate and train images and algorithms. The fields of application are exciting and diverse," says Rentsch, "they range from commercial and standards compliance applications in industry and manufacturing to automotive suppliers and the inspection of microscopic biospecimens in medical laboratories." 

Machine vision yesterday and today

Machine vision refers to industrial applications that use visual systems to guide automated processes. For decades, it has contributed to increasing the efficiency of quality control in various areas of life and industries. However, complex decision-making processes in quality assurance often require advanced techniques that go beyond conventional methods. Deep Learning, as a subfield of artificial intelligence (AI), now also allows more sophisticated checks of facts, units and makes, for example in the case of random arrangement, random shapes or changing assessment criteria, writes IT-Logix. For example, the company supported the Migros subsidiary Micarna in a project for an image classification system based on Microsoft Custom Vision: This is intended to relieve employees at the largest meat, poultry, fish and egg producer of repetitive and physically tiring decision-making processes for the classification of slaughter products (see text below).

Source: IT Logix

Meat processing by means of AI

With artificial intelligence (AI), Migros subsidiary Micarna relieves employees of repetitive and physically tiring decision-making processes for classifying slaughter products. In the cutting plant of Switzerland's largest meat, poultry, fish and egg producer, product identification is supported by an image classification system based on the Custom Vision service from Microsoft's Cognitive Services family, instead of by visual inspection by employees. IT-Logix therefore worked together with Micarna to provide proof of concept. For this purpose, the image classification algorithm provided by Microsoft had to be conditioned to the specific requirements of the company.

"The fact that the system basically works has already been demonstrated together with IT-Logix," says Michael Kott, Head of Project Management at Micarna in Bazenheid. The aim of the project was for employees to receive suggestions for all products from the artificial intelligence and then only have to confirm and post them in the system at the touch of a button. AI is therefore not intended to rationalise people away, but to make their work easier. If the service takes over the physically strenuous visual work of classification, the human can concentrate on confirming the system's suggestion and thus conserve his strength. "Image categorization brings a great relief and also an optimization of the result quality in the long term," says Kott.