Attention: The interview was held in German. This article Is a shortened and translated summary. The original interview can be found here.
The Internet of Things (IoT) has arrived in society and represents a major technology change. Swisscom, for example, is building a complementary network in Switzerland for the IoT. IoT Apps for Raspberry PI and Smartphones are becoming more and more part of our society.
Adapting new technologies such as the IoT for factory automation is no new concept, especially when it offers added value to a reasonable price. IoT can already supply today this added value and as soon as the expected quantities of units are produced, the price will also be right.
The benefit of the IoT is actually that it is based on known technologies, which are deployed in a new way. Data protocols and low-power sensors are already known in their application. New to automation are terms like Cloud or Big Data, that make use of one key element in the IoT: Edge Connectivity.
Edge Connectivity it managed through so-called Edge Gateways, that connect industrial networks such as PROFINET to the cloud. The usage of such a gateway is basically no more complex than the use of a conventional automation device. In fact, it just behaves like an additional participant in the network.
IIot Workshop lab setup - September 2016 at Berner Fachhochschule led by Prof. Max Felser
Industry 4.0 and Industrial Internet of things needs a whole spectrum of skills and competencies. In order to benefit from IIoT, skills and competencies are divided between experts in their respective field.
Process engineers need to know and master the process to be optimized. Automators collect the significant process variables so that these analyses can be carried out by the IT experts.
Therefore, automators must only acquire the competencies, supplement the existing information and make it available to the expert in a suitable form.
Unfortunately, the issue of security is not yet addressed with the necessary attention. Security is merely still neglected in the industry 4.0 by the engineers in automation.
Nowadays, automation systems are often insulated and disconnected from the Internet. If Edge Connectivity is to achieve the full potential, additional competencies are required.
There are currently two main “danger zones”:
1. Data coming from the field to the cloud. This is a matter of confidentiality of the information.
2. Data coming from the cloud into the automation. The question here is: who is responsible for the reliability of this data?
In industrial automation, the protocols OPC UA and MQTT seem to prevail, but comparing the two protocols is like comparing apples and oranges.
MQTT is a very simple transfer protocol, which admittedly meets many of the requirements of the IoT. For the rapid development of prototypes certainly the first choice.
The OPC Foundation sets everything in motion to make OPC UA a fit for all possible applications. With its services and object definitions, it is way more comprehensive and is better adapted to the automation world than MQTT. In addition, the Publish-Subscribe model is taken from the MQTT for OPC UA.
With the TSN (Time-Sensitive Network) as a real-time protocol, interfaces to PROFINET or similar networks will also become superfluous. The development is only at the beginning and it is somewhat early predictions to venture as far as OPC UA other protocols can displace.
In the IIoT workshop at the Bern University of Applied Sciences, participants learn all the basics on how Industrial IoT can be used in factory automation. It is designed for technical professionals, who highly appreciate that we have integrated practical exercises with industrial material directly in class.
One core element in the courses is the Edge Gateway. It is connected to a PROFINET system and behaves like an additional device. The interesting thing in this case, however, is that the Edge Gateway can passively monitor the PROFINET like a bus monitor. This allows it to passively monitor data without any bigger adjustments to an existing configuration and to copy these process values into the cloud for an extended evaluation.
There the appropriate process can be used to monitor the existing process and to optimize any weak points.
With it, ideas from asset management or predictive maintenance can be rapidly implemented. There is also potential in process optimization, where one even assumes a "human in the loop", which means, that the analysis of the collected data allows the operator to set optimized process parameters.
IIot Workshop impressions - September 2016 at Berner Fachhochschule led by Prof. Max Felser
This is certainly a sensible first step until the confidence in these technologies has grown. Until then I personally still lack the clear "best cases": how to do this or that. Today's solutions leave everything open, you do not want to block an option. A functional pattern has been implemented quickly. However, in order to design a secure application, the user is still very much dependent on his personal competences.
1. In automation, we need evolution and not revolution. The constant development has proven itself.
2. Today IoT is where the automation networks were 10 years ago: protocols are there but the data-interrelated data structures are not yet in sight. Here one can safely learn from industrial networks.
3. New IoT paradigms bring new paradigms, which are better adapted to modern security concepts. Here the existing industrial solutions have to improve or they will lose their importance in the long term.