How Digital Twins are revolutionizing Process Industries

Right now, we are at the beginning of a new era in the industrial revolution, which is the most environmentally friendly so far. Now it is all about Smart Factory, autonomous systems, and networks. A large piece of all these innovations is a Digital Twin.
November 24, 2021
Get smarter

Digital Twin

Digital Twin is a great hype in digitalising technology. There are great chances that you have already heard about it. But it is also probable that you didn’t fully grasp the idea of Digital Twin and its purpose. Don’t worry! We are here to help you find answers to your question. We will explain where the idea comes from, what Digital Twin is, and how it’s made. We will demonstrate where this technology can be used and some existing applications. Continue reading to become familiar with Digital Twin and get inspired for action.

Industrial Revolution in time

A significant turning point in history was made during the Industrial Revolution with the transition from hand production methods to machines, new chemical manufacturing, and intense use of iron production processes. This gave rise to the development of machine tools and the growth of mechanisation practices at factories that involved steam and water power.

Industrial Revolution in time

The next step in revolutionising the industry was focused on mass production with assembly lines involvement. At this stage, electricity started to be used for powering machines at the factories.This was followed by further industry improvement when logical components (computers) were implemented for processes automation.

Right now, we are at the beginning of a new era in the industrial revolution, which is the most environmentally friendly so far. Now it is all about Smart Factory, autonomous systems, and networks. And this is the stage when the innovations greatly rely on Digital Twin implementation.

Chemical Industry

The dawn of the chemical industry matches with the beginning of the Industrial Revolution. Like other industries, it also underwent changes in manufacturing processes over the years.

However, now more than ever chemical industry requires drastic improvements to comply with the increasing demand for environmentally friendly manufacturing with minimum waste. It is not the most straightforward job to find alternatives for raw materials, chemical reactions, and refining methods. Nevertheless, the chemical industry will turn the manufacturing wheel in the desired direction by incorporating new trending technologies at different processing stages, including Digital Twin.

"The European chemical industry welcomes the Council’s Conclusions which recognise the chemicals industry’s critical role for advancing the green and digital twin transition in the development of a sustainable and competitive European industry."

Marco Mensink, Cefic Director General, 15/03/2021

What is Digital Twin?

You probably already heard about Digital Twin as it generated considerable interest, especially with the rapid development of the Internet of Things (IoT) and 5G technology.

Digital twin is a virtual/digital representation of a physical entity or system

In a nutshell, Digital Twin is a virtual model that matches a physical object/system/process completely. Having established a real-time connection between physical and digital appearances (Twins), Digital Twin allows you to monitor and control equipment and operations remotely.

Moreover, you get access to perform simulations on the digital copy of the physical entity, test and make predictions for different scenarios and investigate future potentialities. All in all, Digital Twin provides numerous benefits to companies like propelling innovation, improving production (time, quality, efficiency), and speeding up time-to-market.

What do you need to create a Digital Twin?

Let’s start with the necessary components to create a Digital Twin regardless of your application specifics. Of course, these are not the only critical parts, but the final product will not be a Digital Twin without them.

Sensors (IoT)

Those sensors are required to collect all possible information regarding operational behaviours, processes, and data from the surrounding. Overall, sensors will measure/detect/determine different parameters: temperature, humidity, viscosity, vibration, pressure, density, luminescence, etc. In the chemical industry, the number of observable parameters increases because each specific chemical reaction, occurring in the presence of changed surrounding conditions, can end up in a completely different state from the predicted one. Temperature, colour, odour, viscosity, conductivity, density are just a few parameters that can be monitored via various sensors like Soft Sensors, Sensor Fusion, and others.


A well-established communication network guarantees that Digital Twin receives all necessary data securely and reliably from a physical entity. When creating such a connection between a digital world and a chemical product/process, special attention should be paid to link reliability which should be durable, quick, and secure. An excellent example of such communication is a signal from a spectrometer used in our CHAI project, an innovative collaboration between Dotdash, P&G, allnex, and Imec.

A digital platform

All data obtained through the sensors, together with tasks, requirements, standards, business data, etc., is stored on a digital platform. Such data repository also contains all necessary information that Digital Twin uses for data-driven decision-making. For instance, apart from the sensor data in the chemical industry, a considerable space on a digital platform will be assigned to libraries containing all information about chemical entities, reactions, and combinations. Luckily, scientists started working on it some time ago and they created plenty of various databases for multiple usages, such as Chemical Entities of Biological Interest (ChEBI), NIST Chemical Kinetics Database, Open Reaction Database, Spectral Database for Organic Compounds (SDBS), ChemSpider, and many others.

It is worth mentioning that Digital Twin is not simply a data model as it must have relational interaction with a physical model. Thus, properly designed, it should feel like a real product/process/environment. Digital Twin should be able to simulate the models forward with varying degrees of fidelity.

Why do you need Digital Twin?

  • Do you want to know in advance how a hot summer will influence the quality of your product?
  • Do you want to prevent any disaster and not waste your time dealing with unforeseen consequences?
  • Do you want to improve your production without losing your resources (time, money, current production)?
  • Do you want to be on the leading edge in your field?
  • Do you want to solve physical issues faster by detecting them sooner?
The Digital Twin is a living model that drives business outcomes

The current economy demands widening boundaries and mindsets well exceeding commonly used old-fashion traditions. It is essential to think beyond and optimize the existing operations to keep up with the fast-changing and challenging market in the chemical industry. Such challenges can be handled only by employing detailed simulations with good flexibility on constraints, reliable predictions, and a well-realised connection to the actual physical object/process. A Digital Twin is an ideal tool for such tasks as it satisfies all those requirements and provides even further control over your manufacturing.

The flexibility of Digital Twin allows mirroring not only a single process but also a whole system, production, and factory. The autonomy fine-tuning of Digital Twin to specific operational conditions makes it a desirable asset for thriving in manufacturing competition.

Moreover, with a constant inflow of operational data, environment information, and demands, Digital Twin stays updated and helps optimise alternative production scenarios. Digital Twin constantly reaches out to be up to date with the latest environmental and operations data and learns from similar data sets to provide new and original outcomes. Digital Twin can predict and notify about failures and forecast opportunities based on these two processes. This includes overcoming reduced production rates, improving product quality, suggesting alternative raw materials and conditions. Having such prediction in a pocket assures choosing an optimal process scenario for a particular demand.

Suppose a company aims to increase their productivity, efficient energy and resources utilization, and quick reaction to the changes and diversity in the global market. In that case, Digital Twin is a golden key that opens the door to success. Furthermore, companies from the chemical industry are already facing difficulties complying with new and constantly changing regulations that impose severe restrictions regarding raw materials, recycling, pollution, and many other aspects. Using Digital Twin ensures chemical manufacturing flexibility, provides solutions for increased efficiency, proposes unforeseen alternatives to the chemical composition or reactions, and enhances internal chemical products recycling/reusing, resulting in a shorter time to market.

Digital Twin at different levels

Depending on the area of application, Digital Twin can be created based on the level of product manufacturing. Of course, different Digital Twins can be combined to build a more complex environment. Here are just some examples.

Digital Twin of product

Digital Twin of a product involves simulating and validating product properties following individual requirements. At this level, a digital product is tested whether it is stable in time, depending on the environmental conditions; if the chemical reactions stabilized; if it is up to the claimed specifications; whether it complies with voluntary standards; is it up for sector-specific standards, etc. Regardless of the active components (chemical reactions, software, processes, performances, standards), Digital Twin helps to optimise, test, and verify all necessary elements in advance.

Digital Twin of fragment

A digitalisation solution can go as miniature as a small part of an object. Digital Twin can represent even a specific molecule that should be obtained in a particular process. A digitalised copy can test variations in the molecular chain, bonds' strength, cross-linking options, chemical groups' substitution, and many other options. For a "digital molecule," a graph representation is a perfect visual and functional model as it can replicate a molecule structure (graph shape) and the bonds type between the atoms (relationships/edges between the nodes/entities).

Digital Twin of production

The virtual environment for a Digital Twin of production involves numerous aspects, including data collected from the machines and controls sensors as well as from the entire production line. Before the actual operation begins, a Digital Twin model can be used to optimize the production process, avoiding any errors or failure. Following such a strategy, one can save time and cost for customized agile production.

Digital Twin of performance

The Digital Twin of performance requires the input of machines status data, energy consumption, other operational information. By analyzing this data and performing learning on a similar set of data, the built model provides solutions to optimize energy consumption, prevent downtime, recommend changes to improve general performance. In the case of the chemical industry, such Digital Twin can reflect the efficiency of chemical reaction or processes with a further recommendation like increasing temperature and speed of mixing or reducing the volume of added water vapor to the system.

Digital Twin of production line

When creating a Digital Twin of a production line, it is necessary to ensure that information from each segment of this production line is communicated to its digital mirror. Since this model has higher complexity than mentioned above, it might consist of a few Digital Twins representing this line's product or production.

Digital Twin of manufacturing process

The virtual model representing the manufacturing process works with a broader range of data from various sensors. Considering that the manufacturing process involves numerous stages, the amount of obtained data grows significantly. To deal with it, appropriate and reliable digital platforms should be used since apart from collected data, Digital Twin needs to store information used for providing prediction and optimisation solutions.

Digital Twin of factory

The Digital Twin factory is the most complex among all listed here. Usually, it consists of interconnected Digital Twins, which represent all products, processes, production lines, supplies, resources, and other components. One of the challenges, in this case, is good communication between all the sensors and the digitalised model. All involved Digital Twins should be synchronised and reflect the physical factory closely. For example, at a chemical factory, one manufacturing process produces an extensive amount of carbon dioxide. Digital Twin of this factory proposed a scenario where this carbon dioxide can be reused to improve the storage lifetime of another chemical compound that might be spoiled otherwise (which was usual scenario before).

Applications of Digital Twin

Are you not convinced yet that you need a Digital Twin for your product/processes/manufacturing? Check out some of the realized Digital Twin implementations and application ideas.

Training platform

Digital Twin can be used for training new operators or becoming familiar with some unique and rare operations which might occur as unexpected events: start-up, shut-down, slowdown, catalyze, etc. It is a safe place where errors will not lead to safety and environmental incidents, keeping your equipment and production line unharmed.Especially it becomes more critical when talking about roles with higher responsibilities, like chemicals operators.

Always available resource

Independently from the status of your product/process/manufacturing (active, on hold, in preparation, closed down), Digital Twin allows you to continue working on your assets, optimising methods, looking for alternatives, and preparing for speed-up/restart/reconstruction.Having a digitalised copy of an entity enables remote work and opens opportunities for collaboration regardless of physical location. For instance, a chemical operator can use Digital Twin to build an environment that fully represents a complex process. This model can be used at the manufacturing site and remotely without losing any features.

Failure and maintenance prediction

Since Digital Twin possesses all the details about a proper working condition (some golden standard), it can easily spot any anomalies in a product/process/production. Thus, warning signals will alert about the possible failure or maintenance necessity. By being warned about possible changes in production, it is possible to avoid significant losses and mitigate the risk of a collapse.

Discovering new usages

Digital Twin can be useful even in such unpredicted areas as recycling and reusing products, processes, production lines, etc. Without a doubt, performing innovative recycling on a site saves resources, money, time and helps the environment. Moreover, shifting to a new product manufacturing might be much faster as Digital Twin can provide shortcuts or not apparent solutions. For instance, rather than reinventing a new strategy and reordering new materials, Digital Twin can suggest reusing one of the on-site available chemicals or already producing chemical materials alternatively to ordering a new badge of supplementary chemicals.

Smart city

Working with an exact digital copy of a physical city allows analyzing numerous ways of improving the living experience of the population. This includes city planning, operation, monitoring, and management. City Digital Twin provides diagnostics of various elements, starting from the quality of buildings' walls and up to the services like a sewer system, energy distribution, road system, education, health care, and many others. Singapore is an excellent example of such futuristic innovation. A similar attempt is launched for Antwerp city as well.

Some interesting facts about Digital Twin

Digital Twin was first realised in the aerospace industry

In the early days of space exploration, NASA research developed the technology of pairing digital to physical objects as a digital model helped simulate, operate, and analyze physical processes in the aerospace industry. Despite the high needs, budget, and means in space exploration, NASA never regrets using Digital Twin for various requirements and continues using it.

Digital Twin vs simulation

Obviously, simulation and Digital Twin employs digital models to mirror products, processes, systems. However, there are a few significant differences that discriminate against them. First of all, it is about the model dimensions. Typically, simulations focus on smaller entities and cover only a part of a process. At the same time, Digital Twin can comprise not only products or processes but also a production line and even a whole manufacturing process and factory. Secondly, Digital Twin requires bidirectional communication with a physical twin by acquiring data from various sensors and communicating back possible scenarios. On the other hand, simulations do not demand the same amount of information to function. Finally, one of the most significant advantages of Digital Twin on simulations is its ability to study, analyse, explore and predict from various points of view, which is primarily impossible by using simulations only.

World Avatar

The idea of creating a World Digital Twin has been implemented in the World Avatar project. Generally, all aspects of the natural world will be represented in a digital "avatar." One of the aims is to have a closer look at the synergies for resource, energy, and emissions savings by investigating virtual industrial operations.

Fast manufacturing

Only two years were needed to design, simulate, and manufacture a new electric car by a small team of young engineers at Electra Meccanica simply by using the Digital Twin.

Start small

Digital Twin does not have to be made for a whole factory immediately. Digitalization is supposed to be a work-in-progress that evolves with increasing resources (mainly IT) and improving communication between installed sensors and digital models. Implementing digitalization for various components makes transitioning to a complex Digital Twin of a whole machine/process/factory much more manageable.


If you are reading this you probably want to have a Digital Twin of your own. This technology can be compared to flying rockets: some time ago it was impossible to some random person to go to space, but with it changes drastically now. The same can be said about new trends in technology for industry digitalisation. Making analogy between the space rocket and Digital Twin one conclusion can be drawn: the engine for both of them will define the efficiency and success. If in case of space traveling it’s all about the fuel and in Digital Twin it’s all about representing the data. Knowledge Graph (KG) is one of the best solutions due to the following advantages:

  • Knowledge Graph is user friendly and represents data in the way as people would explain it to each other
  • Data in KG retains semantic
  • It is easy to add/find/delete data
  • Search over KG is performed in a similar way as a person would do
  • KG provides extra insights
  • New relationships between the data in KG can be established which would not be obvious when using standard relational databases

Imagine now a Digital Twin rocket that is powered by Knowledge Graph fuel. A work of such entity will be smooth from the launch to interstellar flight. Here is why:

  • To launch a Digital Twin, data has to be uploaded and analysed. Since the KG is very human friendly, it will be no difficulty in starting working with it.
  • Good visualisation and KG explainability will allow to modify and optimise data at any stage without compromizes.
  • A long and thriving performance of a Digital Twin built on the KG is possible due to the flexibility of this technology in terms of adjusting to new requirements.

If this post hasn’t answered all you questions, you are welcome to read a white-paper we published on this topic. If you still have some other questions or concerns, please do contact us

Triggered to built your own Digital Twin?

Digital Twins are upcoming because the right technology became available in the last years. Next to IoT and 5G, the Knowledge graph is one of the enabling technologies for Digital Twin. Where relational databases struggle with capturing the knowledge for your Digital Twin, Knowledge graphs approach data from another angle where expertise, knowledge, and flexibility are central.

Are you interested in building your own Digital Twin? Reach out to and see what the possibilities are. We always start small, deliver fast business value, and scale up.

Download our whitepaper on creating business value with smart data
Short description of whitepaper CTA