Exhibitors & Products
Events & Speakers

B&R, an innovative automation company headquartered in Austria with offices around the world, and also part of the ABB Group since 2017, recently added a powerful feature to its smart camera portfolio. Vision algorithms based on deep learning are expected to not only improve quality, but B&R says they will also increase productivity and reduce scrap at the same time - even making manufacturing processes more flexible in the future. "By working with MVTec, machine builders will have access to the most powerful vision capabilities available on the market," said Andreas Waldl, machine vision expert at B&R. "MVTec is an industry leader in deep learning as well as classic rule-based algorithms - two complementary approaches that both play an essential role in machine automation."

Reading aid with high-speed glasses

The first concrete application to emerge from the collaboration is an optical character recognition (OCR) function based on Deep Learning. Dubbed Deep OCR by B&R, the solution is said to achieve remarkably fast read rates, even for fonts where conventional OCR systems often have to pass. "For many years, we have been working with B&R to develop machine vision solutions for customers that set technological standards, especially for embedded systems," explains Thomas Hopfner, product manager for licenses and interfaces at MVTec. "This claim is also pursued by our latest cooperation. In doing so, customers benefit from the advantages of the deep learning technology Deep OCR without having to sacrifice performance."

Low energy requirements - TOPS performance

The downside of powerful deep learning algorithms is often their power hunger, which can only be met by correspondingly powerful processors. However, despite being implemented as an edge device, the B&R solution should not fall into a critical range when it comes to power consumption because, especially in today's world, machine manufacturers need energy efficiency at least as much as they need computing power. This is also the reason why Hailo was brought on board: Despite the enormous performance of 26 tera operations per second (TOPS), the company's inference accelerators are surprisingly frugal when it comes to energy consumption. "Hailo helps us apply breakthrough speed and intelligence exactly where our customers need it for their fastest and most adaptable manufacturing solutions," said Andreas Waldl.

And Hailo CEO Orr Danon adds, "This kind of collaboration is exactly what we had in mind when we founded Hailo. The new B&R camera leverages Hailo's powerful AI accelerator. Combined with MVTec's state-of-the-art edge devices and advanced machine learning algorithms, we're setting new performance standards that significantly improve productivity and quality in industrial applications."

Video