State-of-the-art intelligent vision not only makes the implementation and use of vision and barcode scanning systems easier than ever before, but also improves quality control, optimizes material and energy consumption and increases traceability and productivity.

A vision system can handle a wide range of applications: It can detect defects on products, check final assembly, count parts, record dimensions and much more. In contrast to manual inspections, machine vision works 24 hours a day with consistent performance and offers greater precision and speed.

Deep learning for complex applications

Based on neural networks, deep learning teaches robots and machines to do what humans take for granted: learn from examples. The use cases for artificial intelligence are becoming increasingly diverse, as shown by speech, text and facial recognition based on deep learning technologies.

In some cases (e.g. qualitative interpretation of a complex scene), human vision may be the best choice, but deep learning can overcome the challenges of judgment-based inspection more effectively than a human inspector or traditional machine vision.

Use of AI for smaller companies

While the successful implementation of deep learning-based image processing projects requires planning, knowledge and dedicated resources, AI-based image analysis is now accessible to smaller companies thanks to the recent development of a technology called edge learning. Edge learning is a subset of deep learning, or artificial intelligence, that allows information to be processed directly on the device. This technology has many advantages.

Image processing systems can check the completeness of goods in logistics, in the pharmaceutical industry, in electronics production, in the manufacture of consumer products or foodstuffs and in many other areas. © Cognex

It is easy to use. No specialized knowledge of machine vision or artificial intelligence is required to set up and use an edge learning-based vision solution. Since the algorithms are pre-trained, edge learning takes less time and only five to ten images to learn how to distinguish unacceptable from acceptable parts.

Key to Industry 4.0

Traceability, which is primarily dictated by regulations in the food and pharmaceutical industries, is also becoming increasingly important in other sectors. It makes it possible to track a part, product or packaging throughout its entire life cycle and is therefore one of the most important issues in the entire supply chain.