Cancel

Top tips

tecnologias de las ciudades inteligentes o smart cities

What are digital twins?

Digital innovation for real processes

Reading time: 7 min

What is a digital twin?

A digital twin is a virtual model that is an exact reflection of a physical object, process, or system. It's used to simulate and study the behavior of a digital product to later efficiently adapt solutions to the real product.

A digital twin is the virtual representation of an object in which real-time data captured by data analytics technologies or sensors is incorporated. It's a type of technology that combines the Internet of Things (IoT), Artificial Intelligence (AI), Machine Learning, andl data analytics.

This booming technology developed in the Fourth Industrial Revolution is in high demand in the new professions of the future. These virtual replicas can analyze real processes, experiment with existing physical objects, and create hypothetical scenarios to predict certain behaviors. We can design virtual twins of anything, from an airplane engine to a wind farm, or even an entire city.

The concept of digital twins was first mentioned in 1991 in the book Mirror World by computer scientist David Gelernter, but it wasn't until 2002 that Michael Grieves from the Florida Institute of Technology applied this concept to manufacturing. He presented its development in a lecture on Product Lifecycle Management (PLM) at a conference of the Society of Manufacturing Engineers.

Several years later, in 2010, NASA used this technology to create digital simulations of spaceships and capsules, and they called it "digital twin" in a roadmap report.

How do digital twins work?

It all begins with a specialist, usually an expert in data science, applied mathematics, or engineering. Their job is to study the object that has be duplicated using a mathematical model that will simulate the original physical object.

They do this using digital twin technology, which works with a software program that collects data from the real world in order to create virtual simulations capable of predicting a product, process, or system's performance. This computer program incorporates different technologies to cross-check massive amounts of data and analyze possible scenarios.

Therefore, digital twins are made up of three main parts: the physical product, the virtual product, and the connections generated between them. The final version of the digital product that needs to be built depends on the quantity of data used to manufacture and keep it up to date.

Advantages of digital twins

The main advantages of digital twin technology are:

Real-time monitoring

They allow us to track a process remotely and in real-time.

Predictive maintenance

They allow us to anticipate problems ahead of time in order to solve them quickly. We can also implement improvements at different levels in the production chain and reduce maintenance costs.

Risk and failure detection

The information and data processed using this technology can help to detect possible errors in the prototype in order to start working on a solution and avoid disrupting the real twin's system.

Resource optimization

Work processes are optimized by improving the production chain, reducing investments in the process, and saving on costs.

Characteristics of the digital twin technology

A digital twin is possible in this industry 4.0 thanks to a series of unique characteristics:

Connectivity: Without it, digital twins wouldn't be possible, as they exist thanks to the Internet of Things (IoT), which allows the physical devices to be connected and communicate.
Standardization: To create a digital representation of a product, process or system, all the data must be dumped into the same computer program. That is, in order to develop a virtual prototype, data is taken from different sources and unified to create one single database, thus, standardizing it.
Reprogrammable and smart: A digital twin can be reprogrammed, even automatically, through sensors on the physical product, Artificial Intelligence, and predictive analytics. This allows us to obtain information to improve the functioning of the real twin, or physical prototype.
Digital trail: Digital twin technology leaves digital footprints so that technical engineers can identify problems. In this way, the error is localized and a solution can be found to prevent it from happening again.
Modularity: Both the digital twin and its physical counterpart can be divided into various parts or layers. This facilitates its manufacturing and monitoring since we can identify exactly which components aren't working well and then repair them to improve the overall process.

Types of digital twins

There are 3 types of digital twins, depending on the production phase of the product:

Digital twin prototype (DTP)

A virtual prototype is created when the actual product doesn't yet exist in order to see in detail how the physical product would work and behave.

Digital twin instance (DTI)

This prototype is created when the real product exists; and in this case, the digital twin is used to run tests in different environments where it can be applied.

Digital twin aggregate (DTA)

Information and data previously gathered from the real product are used to predict and identify its capabilities.

Furthermore, depending on what they simulate, digital twins can be classified by:

An individual looking at a various screens

Process digital twin

Simulates production processes from start to finish to develop a more efficient methodology. It allows us to predict how the process will work and where failures may exist, so we can work on solutions to avoid them. This represents a huge cost savings for companies once they launch the physical process.

System digital twins

A digital representation of the production line can be developed to optimize performance as well as predict and prevent any maintenance that may arise. A more efficient manufacturing process can be designed to detect possible problems and save on costs.

Digital twin aggregate (DTA)

The main objective of these system digital twins is to compile mass amounts of data to optimize all the processes and devise new projects.

Work at Repsol

We're looking for talented people that can help us create value in a sustainable way. Find opportunities that you'll be able to take advantage of to keep growing.

Join our team

Applications for digital twins

A digital twin can be implemented in any type of industrial process. Let's take a look at some examples of digital twins in various sectors:

Energy

Digital twins can efficiently simulate energy resource management, in addition to planning possible scenarios. Furthermore, they can model the performance of an energy plant in real time.

Health

Digital twins can be applied in product and equipment prognoses, comparisons between patient records to find patterns, and can even help in reducing risk in different healthcare and surgical procedures.

Automotive

This technology can be very useful for improving car parts, complete vehicles, production lines, and manufacturing plants. They can even be applied in all phases of the process (design, construction, storage, and more).

Logistics

Applications in this sector are wide-ranging, for example, we can simulate how container fleets are managed, monitor shipping, and even design expansive logistics systems.

An individual looking at a digital control panel

Manufacturing

The production chain of any product can be streamlined to optimize its processes and avoid errors and to reduce manufacturing times and costs.

A real-life example of a digital twin

At Repsol, we' designed an Automated Production Management (APM) and Integrated Flow Model (IFM) digital twin, along with OVS Group, in order to improve the efficiency of assets and optimize production.

The main objective of the APM work flows is to "improve operational efficiency and daily production in an integrated and standardized environment".

By combining information from the workflows with various Repsol data sources, we end up with a centralized tool where operators and engineers can quickly identify, classify, and trace opportunities for optimization.