Vision systems: a continuous evolution

2023-12-20

B&R presents its latest machine vision solutions

Vision solutions B&R include not only Smart Sensors, but also image processing algorithms and a series of illuminators innovative, all completely integrated into the automation system. The architecture designed by B&R therefore places simplicity at the centre, as a direct consequence of integration which, in this context, must also be understood at a software level. In fact, no external tools are needed, but control, parameterization and firmware updates: all elements integrated into the Automation Studio design environment.

Increasingly high-performance industrial vision

The adoption of a new quad-core processor has allowed B&R to significantly accelerate the execution of all types of machine vision applications. A technical solution that helps the Smart Sensor complete vision tasks up to four times faster. The image processing algorithms also benefit from the quad-core processor of the new Smart Sensor. Furthermore, the increase in computing power allows the cycle time of the corresponding applications to be halved. With the latest Steady version of the HALCON library, B&R’s machine vision system now features a just-in-time (JIT) compiler that generates executable machine code when the application is loaded.

Deep learning between computing power and energy efficiency

B&R is also strengthening its portfolio with deep learning capabilities. In fact, an interesting three-way collaboration was born with the vision software specialist MVTec and the AI ​​processor specialist Hailo, with the aim of making the B&R camera a powerful and efficient edge device. Deep learning-based computer vision algorithms are opening up new and exciting ways to improve quality, increase productivity and prevent waste, making production more flexible.

A first product developed under the new collaboration is an optical character recognition function based on deep learning, Deep OCR. This is a feature capable of remarkably high reading speeds, even on fonts that are otherwise difficult to recognize. Deep OCR also allows you not to sacrifice performance, although in general high-performance deep learning algorithms require a powerful processor.

The implementation of the camera as an edge device has energy consumption as a critical point. Today’s machine builders not only need computing speed, but also energy efficiency, and that’s where Hailo comes in: in addition to world-class performance of 26 Tera-operations per second (TOPS), the The company’s inferences also require remarkably low power consumption. In summary, thanks to B&R’s technology combined with the know-how of Hailo and MVTech, productions become faster and more adaptable.