How to benefit from a digital twin – real-life cases

Valerius Yläjoki


Valerius Yläjoki
B.Sc of Economics

Digital twin technology can bring great benefits to manufacturing or processing industries, starting with saving resources. You can already find several documented cases of smaller successful adoptions of a digital twin, and larger scale utilization looms on the horizon. In the ever-growing need for greener solutions and resource-saving adaptations, digital twins will certainly find their place in industry.

The digital twin has been named as one of the driving forces in the industry 4.0 transition. It is no wonder, because its benefits range from saving resources by optimizing processes, to improving the training of personnel. Thus, an understanding of the concept is crucial to success in the changing business environment.

However, you could ask what exactly is a digital twin? One way to think of it is as a counterpart to a physical world machine or cog, or even to a whole factory with its workers. The key aspect of a digital twin is the bidirectional data exchange between the twins, where data flows between the physical and digital twin in real time. Another key element is the lifecycle management of the whole interaction between the twins.

Next, several real-life cases are presented, where the adoption of a digital twin has been documented and the benefits of the implementation listed.

Case 1 – The potential of a digital twin in the district heating of cities

A digital twin provides value, for example, in the simulation of district heating services by utilizing the vast amounts of data the physical network creates and has created in the past. By simulating the future needs of the network’s heating power using the digital twin, models can be created for foreseeable needs.

In the simulation based on the data from the physical twin’s history, a digital twin provides a platform where the simulation can happen. By getting knowledge from the simulation, the physical district heating network’s load can be optimized. This optimization saves resources in the long run by cutting wasted energy.  

Silo AI has implemented a digital twin-like solution in the district heating services of Helen. This “intelligent solution” provides better estimates of the district heating networks’ future energy needs based on the historical data of the physical district heating network1. Another case where machine learning has been adapted for district heating is Ensense’s software solution. According to the company, their solution has saved hundreds of thousands of euros for their customers, as well as natural resources2.

If a digital twin were built as an exact digital counterpart of the physical network for district heating, it could be used to simulate the physical network in all its parts. The simulation would consist of the whole network’s status, down to even the tiniest single apartment inside a block of flats. All this simulation could be done with the digital twin.

Case 2 – A digital twin in optimizing plant maintenance

Plant maintenance, also called industrial maintenance, is a process in which the goal is to reduce breakdowns, promote reliability, and increase uptime in the assets of the plant. One of the strategies in plant maintenance is called reliability-centered maintenance (RCM). RCM requires a lot of information on the assets, and especially on the critical equipment that is necessary for the plant to continue its daily operations. The aim is to allocate resources more efficiently depending on the needs of the plant.3

Critical parts may be serviced every three years, regardless of their performance or condition. The plant must keep operations ongoing, and these parts cannot fail or break during operation. If a critical part fails or breaks down suddenly, it will lead to long operational downtimes, which come at an enormous cost.  

To benefit from a digital twin, these crucial parts could be simulated based on their historical data, and the parts’ vibrations4 could be monitored via different sensors and probes in the process, called condition monitoring, which produces information on parts’ movement and patterns. By having the data about the behavior of the parts, it is possible to simulate the parts’ lifecycle in the digital twin. The data could indicate, for example, that critical parts have a longer lifecycle than the three years after which they are serviced.

Based on this finding, the plant could service these crucial parts at the corresponding time, when they are in actual need of service. By prolonging the life of the parts, maintenance costs can be cut, and the plant can operate longer without a service or a halt in production. This also saves resources and the environment in the long term.

Case 3 – Using a digital twin as a learning tool

If you have a digital counterpart to the actual machine, you can multiply the machine’s digital twin to your liking in virtual reality. Having the same effect that you see in the movie Matrix, you can use virtual reality for training in the operation and servicing of the digital twin’s real-life counterpart.

Think of a complex physical factory machine, for which only one person can be physically present in the maintenance environment. This environment can be easily recreated via virtual reality in a completely different location from the factory where the machine itself operates. The key benefit is having all the digital twin’s elements in virtual reality, where you can simulate the physical twin wherever you need to, in whatever scenario you want to.

Elomatic has produced for Valmet the Virtual Mill5, which is a digital design twin of the customer’s machinery and surrounding facilities. It can be used for training operators and maintenance personnel before start-ups and major shutdowns.

Wider adoption of digital twins looms on the horizon

You can see many of the advantages that a digital twin creates. The described case examples are just the start. However, many challenges remain that are preventing the wider adoption of digital twins. Data, privacy, security, trust, expectations, and IT infrastructure are listed as challenges for both data analysis and the industrial Internet of Things6.

To implement a properly functioning digital twin that truly provides value, it is crucial to understand the different factors at play in the model and in the environment in which the digital twin is applied. The benefits must also outweigh the costs of the implementation of the digital twin for the investment to pay for itself.

The fourth industrial revolution is happening now, and even though the technology has not yet been adopted in the everyday manufacturing and processing industries, it surely will find its footing in the future. Currently, there are many documented cases of smaller successful adoptions of a digital twin in practice, but multilayered digital twins for whole manufacturing facilities and large-scale adaptations are lacking. Domain knowledge has been mentioned as one of the reasons for this.6

For now, the digital twin is gaining traction in the academic world, and more and more companies across various industries and domains are starting to see its potential as a component in their processes. Since the digital twin is all about saving resources and optimizing current operations, it will be relevant in the future to come. Even though the digital twin is still in its infancy, when more solutions start to emerge, it will certainly start being more widely adopted.

Definition of digital twin

Originally, the term “digital twin” was used in a presentation by Dr Michael Grieves in a product-lifecycle management conference held in 2004 at the University of Michigan. Since then, the term has been characterized in multiple different ways, both in academia and in industry. The easiest and simplest definition in research and industry was found to be the following:

“A Digital Twin is a virtual dynamic representation of a physical system, which is connected to it over the entire lifecycle for bidirectional data exchange.”7


  1. Tekniikka & Talous, (2020). Tekoäly suitsii hukkakaukolämpöä – ennusteet paranevat jopa yli 30 %. https://www.tekniikkatalous.fi/uutiset/tekoaly-suitsii-hukkakaukolampoa-ennusteet-paranevat-jopa-yli-30-/e08b625e-5a0b-4c9e-b5cf-6257e083fe03
  2. Tekniikka & Talous, (2022). Säästö jopa satojatuhansia euroja vuodessa – Suomessa kehitettiin uusi ratkaisu kaukolämmön energiatehokkuuden parantamiseen. https://www.tekniikkatalous.fi/uutiset/saasto-jopa-satojatuhansia-euroja-vuodessa-suomessa-kehitettiin-uusi-ratkaisu-kaukolammon-energiatehokkuuden-parantamiseen/
  3. Upkeep, Industrial Maintenance | What is Industrial Maintenance, https://www.upkeep.com/learning/industrial-maintenance
  4. Valmet Oy, Portable condition monitoring, https://www.valmet.com/automation/asset-performance-management/condition-monitoring/portable/
  5. Elomatic Oy, (2021). Immersive virtual learning experiences! https://www.youtube.com/watch?v=uE1XmuwrNGo
  6. Fuller, A., Fan, Z., Day, C., & Barlow, C. (2020). Digital Twin: Enabling Technologies, Challenges and Open Research. IEEE Access, 8, 108952–108971. https://doi.org/10.1109/access.2020.2998358
  7. Trauer, J., Schweigert-Recksiek, S., Engel, C., Spreitzer, K., & Zimmermann, M. (2020). WHAT IS A DIGITAL TWIN? – DEFINITIONS AND INSIGHTS FROM AN INDUSTRIAL CASE STUDY IN TECHNICAL PRODUCT DEVELOPMENT. Proceedings of the Design Society: DESIGN Conference, 1, 757–766. https://doi.org/10.1017/dsd.2020.15