The digital twin concept originated in the early 1990s, so it is not new at all. In the nearly 30 years of its existence, several definitions of what digital can be were created.

‍

While for some it is purely data stored somewhere in a database or in a file system, for others it is a BIM model (Building Information Management), for another, it is a simulation software of production processes, for us, it is live data collection and 3D visualization. Which concept is "correct" is difficult to say because it depends a lot on purpose and uses - one concept makes better sense than another.

‍

In any case, everyone agrees that it is a digital copy of any object, whatever it is - a city, a building, a machine, or the human heart.

The idea that digital twins are data about a given object is tempting by its simplification, but in that case, the digital twins would be any ERP system database, cash register system, or JSON file that would describe this object.

‍

So this is a possible interpretation, but the world is already further than to be limited to just data. This data can contain all the important information about a given object, but the human brain is not a machine, and our ability to interpret the data is given, among other things, by their visualization.

‍

So if we give someone Excel with all the data, which is 20 MB, they got a "digital twin", but we didn't help them much - the data needs to be visualized in a usable way.

Construction companies are slowly but surely moving from 2D building planning to various BIM systems, in which it is possible to plan a building in 3D with all the details. From the walls, water, gas, heating, air conditioning, cabling, to the last socket or sprinkler of the fire protection system.

‍

Architects and engineers, therefore, have tools at their disposal, in which they will prepare a complete set of documentation for the building, which is used during the entire construction to control, plan, etc. However, as is the case in life, the changes that occur on the construction site are often no longer applied to the BIM model, and then it cannot be relied on. Of course, this is not a fault of the system, but it must be borne in mind that at some point the model will begin to differ from reality.

For a long time, the industry was built on the premise that engineers plan the process and it will go according to plan.There is hardly anything further from the truth. The process, even the best one, doesn't go according to plan - and it doesn't matter if it's done by machines or by people.While machines have problems with the variation of the material or the calibration of parameters, people are not better off - fatigue, attendance, illness, discipline, or overall low motivation. So the simulation will help in the preparation, but the twin of reality is not.

It is common for companies to have data in different information systems, as until recently no one thought it would be reasonable to have them all under one roof - but what does accounting have in common with logistics costs, waste management, and the quality of the press, right?Even if they thought about it, the cost of such solutions and their complex implementation could be a problem. Buying a single system that solves everything from the purchase of materials, through production, wages, warehouse, trade to shipping is not easy.Today, however, the situation is different and for real-time management decisions, it is necessary to really get the data into one system and adequately visualize it.

‍

To talk about a real digital twin, it is necessary to display all the data in real-time. The possibilities of integration today and a few years ago cannot be compared - data can be obtained not only from existing process information systems (stocks, OEE of machines and equipment, orders, or deliveries), but also directly from PLC (OK / NOK pieces, various alerts ), or various IoT devices - sensors monitoring the environment, temperature, humidity, noise, air quality, etc.A significant part consists of location data from various RTLS (real-time location services) solutions based on Bluetooth, artificial intelligence, RFID, or UWB.

‍

A separate chapter is the use of artificial intelligence as a data source for the digital twin. Artificial intelligence is not only an analysis of numerical data and an estimate of future development but also e.g. image analysis. It is possible to read data about various processes from the image - about quality, volume, quantity, etc. This data is no different from data produced by other systems.

Most of the digital twins mentioned at the beginning offer real-time visualization on desktop computers - mobile devices have not been thought of much, with the exception of some BIM systems. However, a smartphone today is a device that offers a large enough workspace and fast internet, so the manager should not be limited to sitting at a computer.

‍

The digital twin definitely belongs to the mobile, so that the user has data available no matter where he is. What people think exists only in a movie, can be used today - the technologies for that are available. If a company has several operations, all of them can be digitized and the manager just switches which operation he wants to see. And it doesn't matter if it's in Austria or in China.Efforts to standardize the collection and visualization of such data can be seen in many global companies, as they struggle with the fragmentation of systems and need to unify the data.

‍

Although I focus a bit more on the manufacturing sector in this article, digital twins will also find use in facility management - manufacturing companies also have offices.The collection and visualization of environmental data is a topic that is of growing importance today. Returning to the offices is problematic for many companies - they need to measure their use or the quality of the environment, and without a digital twin, data collection is complex and often handled only manually.

‍

The evaluation of environmental data helps the building manager and owner with energy management, heating or cooling, as well as companies using the premises to ensure the best possible working environment for their employees. Thanks to such data, companies are able to use their space efficiently - they know how many people are actually in the space and according to that they can assess how much space they really need and do not rent space, which is then unused.

‍

Live data on remote locations have become even more important today than they were a few months ago. Whereas a year ago various audits were held regularly around the world, today there is a problem with that. And not only because of the closed borders but also because of the time and travel costs.

‍

If the company has dozens of plants, it is of immense benefit to it when it sees the production, logistics, or warehouse running in real-time on a mobile device. Suddenly, it is not necessary to travel thousands of kilometers to find out what is happening in a given factory right now. It is no longer necessary to look at 2D layouts, which the management does not have to know but sees a realistic 3D model with all the data from the factory anywhere in the world.

‍

However, the vision of digital twins does not end with 3D visualization. A twin can't just be a face, it has to offer more. Once the data is under one roof, it is possible to start looking for relationships between them that are not normally visible. Here again, artificial intelligence has the space, with which it is possible to search for these relations, name them, and then take actions that will help the company's operations go further. This is where the future should go.

‍

Collection - analysis - prediction.

‍