Mr. Hauser, where does the trend towards open source solutions come from?

Hauser: The driver for open standards and the trend towards open source software is clearly the issue of time-to-market. And the consumer market has shown us how it's done: open standards such as TCP/IP, open source Linux and the Raspberry Pi have been in use there for some time. Now this trend is simply spilling over more and more into automation technology. And in all three disciplines: Hardware, communication technology and software.

How do you see this in terms of hardware?

Hauser: Industrial computers and edge computers with open standards and based on X86 and ARM technology have been in use for a long time. What is relatively new is that we are now also seeing increasing demand for industrial Raspberry Pi platforms. In 2018, we developed and launched the first fully industrial-grade single-board computer for series production. Today, we can offer a complete control family on the powerful Raspberry Pi 3B+ single board computer from Broadcom with a different number of digital and analog channels. Raspberry PI, the best-known Linux computer in the world, is now absolutely suitable for industrial use.

The special charm here is that all the software from the Raspberry Pi community can be used on the aforementioned industrial hardware. Young engineers in particular prefer programming languages and software tools such as Python, Node-RED, Scratch, JavaScript, Java and Ruby to the languages we support, such as C, C++, C# and Codesys.

The compact Kontron AI platform consists of an M.2 module with the Google Coral accelerator.

© Kontron

And in communications technology?

Take a look at the emerging TSN standard: TSN as a manufacturer-independent standard is becoming increasingly relevant for Industry 4.0 in order to implement real-time applications. The drivers for the real-time functionality and the functionality for safety-critical applications are based on the open source Fedora Linux 5.4.

"The topic of open source is now spilling over with full force from the consumer market!"

And OPC UA is making enormous progress in the direction of open source with the activities surrounding OSADL. An open62541-based demo application 'OPC-UA pub/sub over TSN' helps to establish data transmission based on the open source software. An updated NETCONF daemon - also based on open source - ensures centralized management and enables the configuration and analysis of TSN networks.

And last but not least, the software?

The topic of open source also applies quite strikingly to complex artificial intelligence with machine learning and deep learning. We already support Intel's neural chips and are presenting an industrial AI platform with Google Coral at SPS connect. Google Coral is a combination of hardware and software for building intelligent devices with fast inference-capable neural networks. At the heart of Coral is the Edge TPU coprocessor, which we use with the NXP i.MX8 processor and support under Linux.

You can download all applications from Tensorflow, develop your own applications for neural networks and deep learning for object recognition and classification, as well as quality inspection of objects. TensorFlow is an end-to-end open source platform for machine learning. It has a comprehensive, flexible ecosystem of tools, libraries and community resources that enable researchers to advance the state of the art in machine learning and developers to easily build and deploy ML-based applications.