From November 27 to 29, the SPS IPC Drives trade fair for automation technology will take place in Nuremberg, Germany, in 17 exhibition halls, this time under the motto “Smart and Digital Automation”. Digitization also stands for a stronger integration of IT systems such as image processing into industrial automation, combined with technical requirements for the compactness of the devices, which should be equipped with decentralized intelligence as well as image and signal processing in real time. This is also the background of new standards for industry 4.0 / IIoT such as the OPC UA companion specifications, including “OPC UA Vision” for image processing.

The image processing solutions presented by us in Hall 10.1, Booth 318 are designed for such a networked production environment: Image and signal processing on FPGA processors guarantee high speeds and data throughput at lowest latencies and low thermal power, available as programmable image acquisition and processing boards or (embedded) components such as cameras and vision sensors.
Image processing applications in the 2D, 3D, multi- and hyperspectral field as well as deep learning with neural networks (CNN) can be implemented in a short time using our graphical development environment VisualApplets via data flow diagrams on FPGA processors also of embedded devices and be used for inline inspection, robotics, quality and inspection tasks, condition monitoring and machine control.

Image Processing in Automation

We will be exhibiting the following factory floor solutions at the fair:

  • High-speed automation solution: LightBridge 2 as a small external image acquisition and processing device, which is integrated into the production and from there transfers the result images to (remote) evaluation computers via ThunderboltTM 2 technology
  • High-speed development solution: With VisualApplets and the extension VisualApplets Embedder, (embedded) vision devices can easily and quickly be programmed with decentralized image processing intelligence (edge computing) – also for deep learning
  • High-speed application solution: For deep learning applications, VisualApplets takes over pre-trained deep neural networks and their weights and integrates them as a CNN operator into the application design, which can be combined with additional image preprocessing, postprocessing, and signal processing. Users simply continue to use their existing image processing system.

Image processing hardware for applications in the automation

Deep learning is entering manufacturing environments

At our booth, we will show a demonstrator to illustrate how new technologies such as deep learning can be used to realize high-speed image processing for complex inspection tasks in production environments – with the lowest latencies and high accuracy. You can get a first impression of this in the video (inspection of defects on reflective metallic surfaces).

Panel Discussion on VDMA Forum
Our managing director Dr. Klaus-Henning Noffz as Chairman of the VDMA Machine Vision group takes part in the panel discussion on “PLC & image processing – what’s next” in German language, together with the companies B&R, Beckhoff, Balluff, MVTec and Stemmer. Presented by Dr. Peter Ebert, Chief Editor inVision magazine.

Thursday November 27, 2018, 11 a.m. to 12 p.m.
Hall 5, Booth 240

SPS IPC Drives 2018
November 27-29, 2018 in Nuremberg
Hall 10.1, Booth 318
Exhibition Center Nuremberg