Many companies and plant manufacturers are aware of the benefits of the Industrial Internet of Things (IIoT) and are increasingly making use of them. The focus here is primarily on networking production and systems with one another and simple data exchange. Cloud-based systems for the remote maintenance of systems and machines create system transparency; status and capacity utilization are always in view. Global access to relevant data and evaluations is possible via the internet, maintenance cycles can be planned and plant utilization or individual components can be checked remotely. This saves time and travel costs, which is a clear advantage not only in cases such as during the corona-related lockdown. Cloud-based operating data, which can be easily linked with other information, opens up new application concepts and business models. For example, a manufacturer can sell a pump capacity instead of a pump.

For this to work, individual, task-specific data storage and evaluation is essential. 'MindSphere' is often used as a cloud platform for this, which stores operating data and makes it accessible via various applications (MindApps). However, in order to make well-founded decisions based on the collected data, it is important to map the plant structure well in the cloud platform. This mapping of the plant structure must be simple and efficient, especially for larger systems, in order to allow economical creation and maintenance. With the 'Visual Hierachy Modeler' (VHM) for Mind-Sphere, Codewerk offers an app that improves the options for mapping the digital twin.
for mapping the digital twin.

The basics of the industrial cloud

The 'VHM app' from Codewerk relies on the visual representation of the digital twin in a tree structure to quickly capture system structures.

© Codewerk

The basis for all IIoT applications is the data from the system. That sounds simple - but for larger systems, huge amounts of data from different levels of the automation pyramid may have to be collected, the relevant data from this stream selected locally using an edge gateway, for example, and only then transferred to the cloud and stored.

In practice, automation technology has evolved structures in mostly heterogeneous system landscapes because the systems consist of different subsystems. Clear structuring of the data is therefore of immense importance here. In addition, the resulting digital twin of the system must be easy to change, as the data structures must be continuously adapted to changes in the system or new requirements after the initial modeling.

Transparent plant structure

A data agent speeds up the modeling process by enabling quick variable selection via drag-and-drop.

© Codewerk

Digital natives are entering professional life and naturally bring their expectations of usability and user-friendliness of tools with them. The VHM app therefore relies on the visual representation of the digital twin. A clear hierarchy in a tree structure from top to bottom provides a framework that is easy to grasp. This division from large to small accommodates human thinking and thus helps with orientation and navigation within the data structure.

Hierarchical tree structures are widely used in automation technology and represent systems from a wide range of industries: Location, plant, sub-plant, system, component et cetera down to the sensor, which may contain the data structures for the measured values. The plant mapping procedure therefore follows the top-to-bottom principle. The hierarchically structured modeling of the VHM app makes work easier, as plant operators and plant engineers can easily find their way around in a familiar environment.
The hierarchical visualization allows intuitive and fast work. In addition, data structures that have already been created can be easily moved via import and export interfaces in order to adapt them for similar tasks and use them again.

The app not only saves time when copying previously created structures, the largely automated modeling is considerably faster than the previous manual creation. This approach offers time savings of 30 to 40 % compared to manual modeling, especially when mapping large systems.

A data agent currently speeds up the modeling process by allowing variables to be selected quickly using drag-and-drop. In future, the app will further expand the principle of import and drag-and-drop and also use new standards. Standardized data types - so-called asset and aspect types - can then be selected directly from a catalog. Assets in MindSphere are the digital representatives of a machine or component (pump, motor, etc.) with different aspects, which in turn group related variables (example: 'energy consumption' with the variables 'power', 'current' and 'voltage').

The basis for such catalogs are standards such as MTP, Umati or eCl@ass. In future, Umati and MTP will also allow connectivity information and the data model to be taken from the respective device or module description. This further shortens the engineering time to the finished data model. An extension to incorporate project and system-specific formats is also being planned. The 'Visual Hierachy Modeler' will then largely take over all tasks; the project engineer simply specifies where and how which data should be displayed and is relieved of the routine work by the app.

Concept confirmed in practice

Samuel Wenz is Managing Director of Codewerk in Karlsruhe.

© Codewerk

The benefits of the app are already being used in practice by over 50 users - from providers of security solutions and services for smart cities, administration and public transport to classic automation systems in the beverage industry, for example.

In the latter case, a demo system consists of a mixer, filling system, labeler and conveyor belts from the so-called assets. Within each asset, there is in turn a wealth of functions, the so-called aspects. In the case of the 'conveyor belt' asset, for example, these are the aspects 'motor 1', 'motor 2', 'light barrier' and so on. This quickly results in a wealth of data for the digital mapping of a system with a large number of assets and the aspects they contain, which only allows a good overview when organized. Typically, there are hundreds of aspects in such a plant. If, as in beverage production, several largely identical processing lines are running in parallel, entire assets for a plant as well as aspects within an asset can be quickly duplicated using copy and paste.

This is also helpful if entire plants have a similar structure. Minor, individual deviations in individual parameters in the assets can then be adjusted manually.