Asset administration shells and digital twins: significant elements for tomorrow’s mechanical engineering

Interview with
Dr Volker Franke, Managing Director HARTING Applied Technologies

Dr Volker Franke is Managing Director at HARTING Applied Technologies, the business unit for tool and special machine construction. Dr Franke and his team are potential users of the asset administration shell and the digital twin at HARTING. The digital twin of various components is used to generate a higher digital twin of machines, systems and plants. He spoke to about practical experiences and future prospects. Dr Franke, what is your role as a user of the digital twin and what are the challenges you are facing?

Dr V. Franke: To begin with, it is important to distinguish between the actual engineering process for machines and plants on the one hand and the application phase of a newly created product on the other. We are dealing with various challenges in engineering: the complexity of solutions is on the rise, because, for example, more and more subsystems have to be integrated into a machine - test systems and connections to traceability systems, applications for providing and processing energy data or also predictive maintenance approaches in the future. The world of mechanical engineering has become a lot more complex?

Dr V. Franke: Oh yes, that's definitely the case. More and more IT systems are being introduced, which have to be taken into account right at the outset of the engineering process. At the same time, we have to face the situation that the specifications for the problem to be solved are not clearly described and only emerge over time instead. So we are dealing here with a major share of agility in the development process. In addition, we have to consider both the skills and the availability of developers. This means that besides the availability of staff in general, it is becoming increasingly difficult to find precisely those people who have the right skills to come to terms with such complex tasks. How are you addressing the challenge of automating the engineering process?

Dr V. Franke: We are dealing very intensively with the topic of "Systems Engineering". Today, it is no longer sufficient to illuminate the complexity from a mechanical or electrical point of view. Instead, one must take on the perspective from a systemic point of view. Here, systems engineering provides methods and artefacts that help to define the system, develop it successively and verify its functions when bringing things to a close. And it is precisely these artefacts that in turn help me to remain agile, because they enable me implement further developments in individual stages. And finally, in terms of automation, the digital twin and the asset administration shell come into play. The asset administration shell describes a component according to a standardised procedure in such a way that I can align the requirements from my solution with the provided capabilities of the component and thereby automatically perform the selection, example. What do you expect from the management shell of a connector?

Dr V. Franke: In mechanical engineering, the connector is not the first thing we think of, for example, but instead an electrical axis consisting of a linear unit, a motor and a frequency converter. In terms of their performance parameters, these elements must be designed to fulfil the task of moving a weight of ten kilos by 500 millimetres within 0.5 seconds, for example. This in turn determines the dimensioning of the components that are required. At this point, the connector also plays a role, because it offers the functionality of being able to connect this motor. My demands made on the asset administration shell of a connector are in the presentation of the performance parameters: what current can it transmit, does it enable internet-based communication or does it have a special format such as single-pair Ethernet, etc.? I expect the Digital Twin to provide me with information to include in a schematic design. But I would also expect to be given information so that I can easily integrate him in the mechanical design. And I would also like be provided with installation and operating instructions, for example, or to be given information about the CO2 footprint associated with the manufacturing. In a nutshell: the demands I make on the digital twin primarily relate to the engineering of the machine. However, the asset administration shell can also provide important information for the utilisation phase. For example, metadata for integrating the individual components into energy management systems or data on materials for disposal. Do you regard the connector as a system component?

Dr V. Franke: No, I still perceive connectors as connectivity components that are then inserted into the management shell of a mechatronic component or module, even if the connector is equipped with additional functionality. Let's assume I have to implement the requirement of high availability in my application and I want to use temperature, vibration and mating cycle monitoring in the connector. These factors will have to be planned and must be mapped in the asset administration shell so that I can select the component accordingly. Without an asset administration shell, I would not be able to find a connector again in the engineering process.

The end user, however, will not see the state of the connector, but an aggregated, consolidated state, for example for a machine module. In the shell model, the overall character can change from shell to shell. In principle, this is comparable to a real construction: When I assemble a machine, I no longer primarily see the screws involved, but rather the entire machine and its function - comparable to an oil painting in which the individual strokes create the overall picture. Where do you see the greatest benefits of the digital twin?

Dr V. Franke: In terms of engineering, I can make the processes much more efficient and flexible, and adapt customer requirements in real time, so to speak. On the programming side, efficiency and even automation are equally important in order to make the potential data stream manageable among the diversity and variety of data. With regard to lifecycle services, it will be possible to offer more services in the future, but at the same time to massively reduce the efforts these offerings will require.