The annual Automatica exposition last week provided examples of the latest developments in industrial automation, control software, collaborative robots, and machine vision advances.

More than 46,000 attendees and 890 exhibitors gathered in Munich, Germany, for sessions and exhibits around advanced manufacturing, service robotics, assistive technologies, and artificial intelligence.

Robotics Business Review will be looking at announcements from Automatica 2018 around European robotics, government policies, and 3D printing, among other topics. Let’s start with machine vision.

In the first quarter of 2018, sales of machine vision components and systems in North America increased 19% year over year to $709 million, reported the AIA, or Advancing Vision + Imaging, which is part of the Association for Advancing Automation (A3). Demand for cameras, optics, and software helped propel that strong growth, said the AIA.

Machine vision software makes CPUs competitive with GPUs

MVTec Software GmbH released the newest version of its vision software, HALCON 18.05, which features a pretrained convolutional neural network (CNN) and operates on CPUs. According to MVTec, a standard Intel CPU can reach the performance of a mid-range graphics processing unit (GPU) with a runtime of approximately 2 milliseconds. This enables industrial PCs, which usually do not utilize GPUs, to be used for deep learning-based classification.

Other improvements over the previous version of Halcon include the ability to access Halcon development (HDevelop) procedures in .Net, as well as in C++ and automatic release of unused memory, reducing the risk of memory leaks.

With enhanced edge detection, the software also can more reliably read bar codes, including those with very small line widths and those are slightly blurry.

The quality of the bar-code scans has been verified in accordance with the most recent version of the ISO/IEC 15416 standard.

The new release also offers enhanced surface-based 3D matching to determine the position of objects in 3D space more reliably, making development of 3D applications easier.

Detecting defects gets easier with smarter systems

One of the increasingly important uses of machine vision is for inspection, with automated systems detecting various defects without the need for human intervention. Vitronic showcased various systems for optical surface inspection targeted for the automotive industry as well its VIRO WSI solution, the next generation of weld seam inspection systems.

The company’s inspection systems detect not only small defects during the production process, but they also provide data for process optimizations, opening the door for smart factory solutions.

Vitronic Vinspec products support the inspection of surfaces and are fully automatic, detecting material- and production-induced defects in automotive parts. The advanced internal inspection sensor, a new addition to the Vinspec family, can be used to inspect interior walls of cylinders that are not easily accessible, and can also capture data for different processing planes across the entire surface.

The company’s Vinspec 3D Inline helps companies inspect surfaces and geometric deviations on castings in production cycle as well as the surfaces of small components, such as bearing blocks in engine blocks will also be presented.

All inspection solutions send results from Vitronic products and from external inspections to an integrated database for documentation and archiving purposes. There, it is organized, visualized, and analyzed so it can be used effectively.

Defects identified by the system can be displayed there on the basis of different views of the test part, a feature that significantly speeds up reworking.

In addition, Vitronic provides the data basis for automated rework of defective welded parts through specially developed software algorithms.

The company pointed out that, because of high standards for each part of automotive battery cells, optical systems for welded seam inspection of battery cells are essential for objectively verifying compliance.

PC software makes industrial imaging easier to do

Simple usability without complex systems is increasingly important. Matrix Vision GmbH unveiled its new stand-alone PC version of mvIMPACT Configuration Studio (mvIMPACT-CS), the software core of the mvBlueGEMINI smart camera.

Matrix Vision’s software enables users to start with industrial image processing without any programming knowledge and without any knowledge of image processing thanks to its intuitive and simple operation concept. The mvIMPACT-CS software offers a wide range of tools to help speed up the development of the application.