Digital Twins, keys to a term that you are going to hear a lot from now on

There are several technologies that come together in the so called digital twins or, by its literal translation, digital twins. As their name indicates, they are ultra high fidelity virtual reproductions of a real scene. This can be a factory, a warehouse, or any space that involves operational complexity. And it is this complexity that leads to the creation of the 3D scenario, aimed at improving efficiency in the physical world.

To achieve such high precision and by nature of their use, digital twins are armed with 3D design tools, 360 degree cameras, and digitizing software. All this allows you to create environments ready to experiment on them, without the need to modify real variables. In other words, tests can be carried out for possible improvements and guarantees of the results can be obtained before implementing the changes in the physical world. This reduces the risks of making wrong decisions.

What are digital twins for?

Actually this technology is applied to study systems or objects with high precision. From IBM they put the example of the turbine of a wind turbine. This is equipped with a set of sensors in vital areas for its functionality. From there, devices start generating data. Everything possible is captured in relation to the performance of the turbine, such as the electricity it generates at any time with wind energy, temperature, weather conditions and endless other details.

All this information is sent to an algorithm, which processes it and applies it to a digital copy of the turbine. The virtual model of this machine allows you to run simulations, varying the data at will. In this way you can check the performance it would have in different situations or study possible problems. In the same way, improvements could also be sought to anticipate the deterioration of the piece.

Although, obviously, digital twins can also be a virtual representation of an industrial plant, where people and robots work in the same space. And we must not forget that the technology is equally applied to the study of systems that are difficult to understand if not through simulations. This would be the case of the architecture of a telecommunications network, whose digital twin could be used to simulate cyberattacks and test defense formulas.

Based on a set of technologies

In a virtual model of a warehouse, people and autonomous machinery can move at the same time. They will do so according to the data collected by the sensors that feed the system. It is the ideal scenario for developers to include new robots or change processes. This way they will discover what could happen in the industrial building if they carry out these changes.

But creating such a model is not easy. It requires careful design, but also precise information management. If a test is introduced, such as a new robot in the plant, algorithms are needed that anticipate the movements of the employees, according to the data that we have about them. This extrapolation has to be reliable to apply virtual solutions that are later viable in the physical world.

A mix of technologies is used to create a digital twin. “Different methods are used such as CAD design, video and photography using 360 cameras or scanning with lidar cameras. The use of drones flying around buildings, towers and other tall constructions is also frequent”, points out Giovanni Cetto, CEO of TwoReality, specialized in virtual reality.

All the graphic information captured is dumped onto a platform. “As metaverses are computer designed virtual sites, these 3D images are inserted into those spaces to create the digital constructs in which we enjoy virtual reality experiences, allowing us to visit the spaces inside and out and placing components in the digital construction”, points out Cetto.

The differences between digital twins and simulations

Both digital twins and simulations are digital models that replicate a system. The main difference is in the scale of detail. A simulation usually involves the study of a specific process, while digital twins are more complex. With them you can study multiple processes. Hence, its technology is often associated with the metaverses, due to that element of set and complexity.

Simulations, for example, do not have real time data. However, the comprehensiveness of data collection for digital twins allows have a constant flow of information to feed the model. Intelligence can also be extracted in real time.

What real data is needed

The goal is to make an accurate portrait of reality. Therefore, the more information we have, the better. “Geographical positioning, physical coordinates, height, width and depth measurements taken by cameras and even marks, cracks, cables, connectors and other architectural details are used,” explains the CEO of TwoReality, referring to a model for a industrial plant.

Much of this data is collected thanks to IoT sensors. This technology helps to connect machines and devices, so that they share information. And all that baggage they generate can be transferred to a digital twin. In addition, important data can be obtained from other elements. “It is also possible to use architectural plans and underground conduits to add them to the digital model”, says Cetto. “So utility companies can know what’s below the surface and work efficiently without damaging infrastructure.”

The applications of digital twins

A study published by Markets&Markets estimated the digital twins market at 3.1 billion dollars in 2020. By 2026, the expected jump is such that the analyst firm forecasts a volume of 48,200 million dollars. This would imply a compound annual growth rate of 58%.

These estimates give the technology enormous growth. But the truth is that digital twins bring significant advantages to companies. And lately its development has been simplified and cheaper. “Thanks to them it is easy to build models to present information to work teams. To designers, architects, operations teams and project managers or investors”, details Cetto.

The CEO of TwiReality summarizes the benefits as follows: “With this tool it is possible to create imaginary situations to model emergency reactions, redesigns, time and motion analysis, volumetric analysis and construction progress, among many others. You can even go into the digital models to look inside and interact with the elements.”

They are also related to the new wave of automation that will roll out across the industry. Efficient coexistence between workers and robots will be developed based on tests, but first it will be cooked on digital twin platforms.

IBM puts some other use cases, like in healthcare. One could even make a model of a patient, with the indicators that are available. Thus, before applying a treatment, it could be simulated virtually to have some clue of what the result will be. Urban design will be another application. With a system that represents a city or a neighborhood, urban planning improvements can be proposed and check how the movements vary.

An improvement in efficiency

The main advantage offered by digital twins is the improvement in efficiency. This is how Cetto asserts it: “Virtual reality simulations can be created in which designers and the operations team can move and change elements, such as conveyor belts, warehouses, the direction of work flow, shelves and operator positions, to determine finally, which is the process that has the best performance or the fewest failures”.

It can be seen as a kind of dress rehearsal. But without everything that implies any change in the daily reality of the industrial plant or any other system under study. “Physical tests, which require more time, resources and waste of materials, are not necessary,” Cetto points out. Simulations can be used to solve problems, such as bottlenecks, without too much experimentation. Engineers are able to go more directly to the solution.

Featured Image: Nikotxan