Information Models for the Digital World

  • Post category:PI NEWS
  • Reading time:4 mins read

Do you remember the book “Measuring the World” by Daniel Kehlmann? It tells about how Alexander von Humboldt set out to bring more structure to our view of the Earth – not just qualitatively, but quantitatively as well. The known world was getting larger and larger, discoveries were revealing the world’s rich diversity, and people needed some kind of orientation. Alexander von Humboldt pointed out the connections between climate zones, vegetation, and utilization by people for the first time.

What does this have to do with PROFIBUS and PROFINET International (PI)? The volumes of data being generated and processed in industrial production systems are increasing tremendously. The spectrum of available technologies is also growing. IO-Link and, most recently, omlox have been added to the canon of PI technologies. Edge computing, as well, is making its way into automation, as data from the classic automation pyramid is required in new value-adding applications with minimal engineering effort. Advanced asset management, condition monitoring, and AI-based devices and system analyses require huge volumes of data.

Measuring with sextants, yardsticks, and other instruments of the kind doesn’t work in the digital world. We bring about the needed order in data by relating it to its object of origin and describing this. This is how information models are created. We designate the importance of the data, such as a temperature, and also describe the transducer that provides the temperature.

The first profiles arose at PI in the early 1990s when fieldbus technology was just being created. At the forefront of innovation, sensor and actuator profiles were designed and function block technology implemented in the devices and remote I/O of process technology. The first information models had been created. The PROFIBUS and PROFINET networks and services had to be described as well. The GSD file is the first example of an information model of the communication-based data of a PROFIBUS and PROFINET device.

Information models need a clear concept vocabulary and organizational structures. Another illustration we could draw on from our lives would be 18th-century scientist Carl von Linné. von Linné, who was what we would today call a biologist, defined the plant classification system still in use today. This includes, for example, monocots and dicots, labiates and cruciferous plants, etc. Mr. von Linné also shaped scientific nomenclature. In our technical world, we also have a classification and concept-forming standards like this. PI is working together with ECLASS to bring PI technology together with initiatives important for engineering applications. A white paper describing the connections of information models with definitions from ECLASS will appear soon.

Today, OPC UA is an important trend for providing information models in a machine-readable format. So-called Companion Specifications are the profiles of OPC UA. This is why it’s only a small step for PI to provide OPC UA Companion Specifications for PROFINET and IO-Link technology, energy management, the information model for PA devices (PA-DIM), and safety as information models. All PI technologies are being reviewed with respect to their information models and standardized for OPC UA – and with this, installed PI technologies will once again leverage innovative applications. Finally, we are writing in the book “Digitalization of the industrial world.”

Follow this link to learn more about Information Models

Prof. Dr. Christian Diedrich
PI Quality Committee Chairman
Otto-von-Guericke-Universität Magdeburg
ifak- Institut für Automation und Kommunikation e.V. Magdeburg