What is a digital twin?

With the emergence of Industry 4.0, the digital twin is becoming increasingly important. But what possibilities does the use of the digital twin offer? And what needs to be considered during implementation?

The Digital Twin

Applications and specific examples

The following two practical examples will illustrate where and with what intention digital twins can be used in industry. Let's start with the first example:

The virtual commissioning of production plants

Initial situation: conventional commissioning

There will be no way around physical commissioning in the future. After all, goods will continue to be produced in the production halls in Industry 4.0 and will require the corresponding hardware for manufacture. However, with the establishment of digital twins, the nature of commissioning may change very significantly. This is because the classic commissioning of plants (e.g., mechatronic systems) has a number of inherent disadvantages as well as risks. 

For example, subsequent adjustments to plants are expensive and sometimes very time consuming [5]. Real systems can be damaged and often there is no time for systematic testing of multiple scenarios and components [6]. In many cases, there is little room for optimization and improvement on the plants. Faults are not detected or are detected too late, which can result in high follow-up costs due to maintenance and downtime costs due to plant shutdown.

In addition, non-compliance with technical requirements and specifications during commissioning leads in the worst case to contractual penalties or at least to unhappy customers. These disadvantages are definitely to be avoided from a business point of view.

Solution: Virtual commissioning through the use of digital twins

Virtual commissioning can drastically reduce development times. In addition, this offers the possibility of avoiding production downtimes and malfunctions during operation [7]. In virtual commissioning, digital models of the production plants or machines are created and mapped three-dimensionally. All properties and behavior of the equipment (e.g., electrical, mechanical, thermal, dynamic behavior) are simulated using models and algorithms. This holistic approach ensures the best possible approximation of the system behavior of a plant [8]. Using common simulation tools (e.g. FEM or CFD tools), the components of a complex plant as well as their interaction are tested under different conditions.

This simulation provides early information and insight into possible sources of error during subsequent operation, e.g. malfunctions, anomalies or problems at the plant. Corrections to the 3D model improve the interlocking of the electrics, software and mechanics even before the actual commissioning. One advantage is that critical scenarios and situations can be tested and evaluated virtually - and not on the real plant. This reduces the risk in real operation, lowers costs and results in a reduction of production downtime in the future. The decisive last step is to connect the optimized simulation software with the real control of the plant (PLC) [8].

Predictive maintenance for optimized maintenance.

Initial situation: Too inflexible maintenance and waste of resources

As already mentioned in the first example, machine failures on industrial machines cause very high costs, which often far exceed the cost of the component to be replaced. In this context, it becomes clear that companies should strive to reduce downtime and associated losses through more systematic and predictive maintenance [9]. However, the reality on the operational store floor is usually different:

Many companies still rely on a reactive maintenance strategy. This means that the affected equipment is not serviced until a fault or problem has already occurred [10]. This form of maintenance is reactive because it is unplannable and occurs unexpectedly. At the same time, companies risk a longer and expensive downtime, which is necessary for the repair or replacement of the components.

It is also common practice to maintain technical equipment or buildings at predefined intervals (for example, an inspection every 3 months; a filter change every 2 years ...) [11]. Manufacturers therefore pursue the goal of minimizing the probability of downtimes preventively [12]. The problem here is that machine components are sometimes serviced on the basis of the specified maintenance interval, even though they are still in good condition and fully functional.

Solution: Application of the digital twin for predictive maintenance

The good news is that in this case, too, the digital twin can deliver great added value and raise maintenance in the company to a higher level. We are talking about predictive maintenance, which is the next evolutionary step up from preventive maintenance. Mr. Feldmann of Roland Berger calls predictive maintenance "one of the key innovations of Industrie 4.0" [13].

First, operating parameters and process parameters of the plant must be measured, transmitted and stored by means of sensors [14]. With the help of creating a digital twin of the plant, engineers can intentionally simulate failures in a risk-free environment and thus capture certain failure conditions of the real plants [15]. Different degrees and manifestations of possible machine faults are combined and evaluated. Then, a predictive maintenance algorithm is trained with this simulated dataset in order to later detect faults on the plant in real cases and classify them [15]. 

In this DT example, the digital twin is also used to simulate the consequences of changes to an object and increase planning reliability for companies [16]. AI-based predictive maintenance algorithms are not only able to detect faults as they occur, but also to reliably predict future potential faults and thus forecast optimal maintenance times.

Added value through the use of Digital Twins

As you have seen in the two examples - DT for virtual commissioning and DT for predictive maintenance - a variety of benefits can be generated by using digital twins. Some added values are listed below:

  • Increase in lead times
  • Increase in overall operational efficiency [17]
  • Reduction of machine downtime
  • Increase in productivity
  • Faster and more targeted identification of faults, bottlenecks and error-prone processes
  • Data generation during use: New insights and backgrounds through real-time coupling with real objects; derivation of new business models through deeper customer and process understanding
  • Improved transparency and information
  • Optimization and control of processes [18]
  • Minimization of risks and errors [19]
  • Reduction of dependence on prototyping in product development
  • Recording of current operating states, but also forecasting and prediction of future states and events → e.g., preventive quality assurance (e.g., through predictive maintenance based on generated data as well as AI algorithms)
  • More reliable overall planning

Recommendations for successful implementation

Digital twins can bring significant added value to your business. However, this requires a clear commitment to digital transformation and data-based processes. It is crucial that the digitization strategy in the company is in line with the ambitions in the field of digital twins. In particular, the leadership in the company must prioritize the topic and allocate the necessary financial budget for pilot projects, upgrading to sensor technology, commercial simulation tools, and AI-based services.

Many digitization projects fail due to poor planning and communication. The complexity of the project is also often underestimated [20]. Driving forward, building as well as continuing DT projects is usually more efficient if experts in the field of cloud, software development and IT service providers are involved [21]. Despite all the justified enthusiasm about the many possible uses of digital twins: The digital twin plays to its strengths above all in complex systems [22].

Mr. Hartmaier from IBM Watson IoT advises the motto of first thinking about meaningful use cases instead of acting hastily [23]. It must be analyzed where gaps exist and in which areas the company can benefit in the long term by developing a digital twin.

Decision-makers must first be clear about what goal they want to achieve with DT deployment. Do they want to make their operational processes more efficient or, for example, offer customers data-based services? Initially, it is advisable to implement small processes, gain experience and get the workforce on board. Then, larger processes can be implemented step by step and the use of DT can be scaled in stages [20] DiConneX recommends answering the following questions regarding goal setting and data literacy in the company [20]:

  • What does my company want to achieve with the digital twin?
  • In which areas should the DT be used?
  • What data is needed and in what quality?
  • How does my company collect the data and information, and how do I subsequently evaluate and analyze it?
  • Do I have efficient data management?

Your journey with OHB Digital Services

Does this sound interesting for you and your company?
Then get in touch with us.

1    Rosen, R. et al. (2020): Simulation und digitaler Zwilling im Anlagenlebenszyklus. Virtual Commissioning of Automation Systems, VDI/VDE-Gesellschaft Mess- und Automatisierungstechnik

2    Stark, R. (N. A.): Smarte Fabrik 4.0 – Digitaler Zwilling. Fraunhofer-Institut für Produktionsanlagen und Konstruktionstechnik IPK, Berlin

3    Ramm, S., Wache, H., Dinter, B., Schmidt, S. (2020): Der Kollaborative Digitale Zwilling: Herzstück eines integrierten Gesamtkonzepts. ZWF Zeitschrift fuer Wirtschaftlichen Fabrikbetrieb 115(Special): 94 – 96.

4    Mitache, R. (2018): The Digital Twin Organization: Can Enterprise Architecture Help? BiZZdesign. Online abgerufen am 07.10.2020. https://bizzdesign.com/blog/the-digital-twin-organization-can-enterprise-architecture-help/.

5    Freyer, B. (2019): Warum ohne die Virtuelle Inbetriebnahme heute nichts geht. Machineering. Online abgerufen am 05.10.2020. https://www.machineering.de/blog/wissen/article/warum-ohne-die-virtuelle-inbetriebnahme-heute-nichts-geht/.

6    iT ENGINEERING (2020): Virtuelle Inbetriebnahme (VIBN). Der digitale Zwilling bei der SPS-Programmierung. iT ENGINEERING SOFTWARE INNOVATIONS. Online abgerufen am 06.10.2020. https://ite-si.de/virtuelle-inbetriebnahme/.

7    KUKA (N. A.): Engineering. Experten für Ihre automatisierte Produktion. Online abgerufen am 04.10.2020. https://www.kuka.com/de-de/produkte-leistungen/produktionsanlagen/technologie-consulting/engineering.

8    ISG virtuos (N. A.): Virtuelle Inbetriebnahme (Definition). Online abgerufen am 05.10.2020. https://www.isg-stuttgart.de/de/isg-virtuos/virtuelle-inbetriebnahme.html.

9    Wallner, P. (N. A.): Der Digitale Zwilling als Baustein von Industrie 4.0. Online abgerufen am 06.10.2020. https://www.maschinenmarkt.vogel.de/der-digitale-zwilling-als-baustein-von-industrie-40-a-904923/?p=2.

10 van Dijk, N. (2017): Von der reaktiven zur proaktiven Instandhaltung: wie Wartungspläne überflüssig werden. PLANON. Online abgerufen am 06.10.2020. https://planonsoftware.com/de/ressourcen/blogs/von-der-reaktiven-zur-proaktiven-instandhaltung-wie-wartungsplane-uberflussig-werden/.

11  PLANON (N. A.): Geplante präventive Wartung. PLANON. Online abgerufen am 07.10.2020. https://planonsoftware.com/de/glossar/geplante-praventive-wartung.

12  Günther, J. (2020): Instandhaltung 4.0. So klappt es mit proaktivem Service in der Instandhaltung. Instandhaltung. Online abgerufen am 05.10.2020. https://www.instandhaltung.de/instandhaltung-4-0/so-klappt-es-mit-proaktivem-service-in-der-instandhaltung-316.html.

13  Feldmann, S. (2017): Predictive Maintenance. Roland Berger GmbH, München.

14  NC Fertigung (2020): Wie funktioniert Predictive Maintenance? NC Fertigung. Online abgerufen am 06.10.2020. https://www.nc-fertigung.de/wie-funktioniert-predictive-maintenance.

15  Miller, S. (2019): Predictive Maintenance emit einem digitalen Zwilling. MathWorks. Online abgerufen am 05.10.2020. https://de.mathworks.com/company/newsletters/articles/predictive-maintenance-using-a-digital-twin.html.

16  Klibi, K. (2020): Der digitale Zwilling in der Intralogistik: Planungssicherheit erhöhen, Investitionen absichern. Miebach Consulting Whitepaper. Miebach Consulting, Frankfurt am Main.

17  Scheibe, H.-G. (N. A.): Schritt für Schritt zum virtuellen Prozesszwilling. Neue Chancen für die präzisere Analyse und Gestaltung von Wertschöpfungsnetzwerken. ROI Management Consulting AG.

18  Uhlenkamp, J-F., Hribernik, K. A., Thoben, K.-D. (2020): Wie Digitale Zwillinge Unternehmensgrenzen überwinden: Ein Beitrag zur Gestaltung von Digitalen Zwillingen mit unternehmensübergreifenden Anwendungen im Produktlebenszyklus. ZWF Zeitschrift fuer Wirtschaftlichen Fabrikbetrieb 115: 84-89.

19  Lambertz, B. (2019): Digital Twin. Maintcare. Online abgerufen am 05.10.2020. https://maint-care.de/knowhow/digital-twin/.

20  DiConneX (N. A.): Digitaler Zwilling – Womit fange ich an? DiConneX GmbH. Online abgerufen am 05.10.2020. https://diconnex.com/blog/2019/08/19/digitaler-zwilling-womit-fange-ich-an.

21  Device Insight (2020): Was ein Digital Twin leisten kann und was nicht. Online abgerufen am 06.10.2020. https://www.device-insight.com/was-ein-digital-twin-leisten-kann-und-was-nicht/.

22  elunic (2020): Was ist ein Digitaler Zwilling? elunic AG. Online abgerufen am 04.10.2020. https://www.elunic.com/de/digitaler-zwilling/.

23  Hartmaier, S. (2018): Der digitale Zwilling. Mit kleinen Schritten angefangen. IT & Production ONLINE. Online abgerufen am 06.10.2020. https://www.it-production.com/produktentwicklung/digitaler-zwilling-small-steps/.