What is a digital twin and where is it used?

A Digital Twin is a digital (virtual) model of any physical objects, systems, processes, or human workers, which accurately reproduces the form and actions of the original and is also synchronized with it. A digital twin can be used to simulate the behavior of a physical object under various external conditions, which helps to save time, money, and avoid harm to people and the environment.

A digital twin not only reflects the current state of the object in real time, but also allows to predict its future state. While statistical simulations based on equations have been used for decades to predict the behavior of systems in the world around us, modern digital twins are adding AI, machine learning, and IoT technologies to their inventory, allowing for levels of prediction that cannot be achieved with traditional simulation methods alone.

Types of digital twins

Digital Twin Prototype (DTP). A virtual analogue of a physical object which actually exists in reality. It includes data for a comprehensive characterization of the model, including information on how to create it in real conditions. These are the requirements for production, a three-dimensional model of the object, a description of technological processes and services, requirements for disposal.

Digital Twin Instance (DTI). Data on the description of a physical object. Most often this data contains an annotated three-dimensional model, information on materials and components used in the past and present, information on the processes performed in all time periods, test results, records of repairs performed, operational data received from sensors, monitoring parameters.

Digital Twin Aggregate (DTA) is a computing system of digital twins and real objects that can be controlled from a single data center and exchange data inside.

What tasks do digital twins help to solve?

• Identify a problem or vulnerability before production starts or an object becomes operational.
• Increase business competitiveness and profitability.
• Carry out a test run of a process or production chain quickly and without significant investment.
• Improve efficiency of processes or systems by tracking all potential failures before starting.
• Build long-term forecasts and plan the development of a company or product for years to come.
• Increase customer loyalty through accurate forecasting of demand and consumer qualities of the product.
• Reduce risks, including financial ones, as well as those related to safety for the life and health of personnel.

How is a digital twin created?

Digital twins can be created in a variety of ways:

• IoT-based model
• visualization technologies, such as: holograms, AR and VR
• graphic 3D model
• integrated mathematical models of CAE-systems (Computer-aided engineering, solutions for engineering analysis, calculations, and simulations)

The steps for creating a digital twin are as follows:

Object study. If the object is already in existence, then a prototype map is created, reproducing all the processes and characteristics of the object under various operating conditions.

Modeling a digital copy of the object. A digital 3D model of the object is created, for which mathematical methods of calculation and analysis are used, such as:

• Finite Element Analysis method (FEA) allows to calculate the operational load. It is used to calculate the mechanics of a deformable solid, heat transfer, hydrodynamics, and electrodynamics.
• In the Failure Mode and Effects Analysis (FMEA) models are required to analyze the reliability of systems and identify the most critical steps in production processes.
• Computer-Aided Design/Drafting (CAD) models are used to calculate the external characteristics and structure of objects, materials, and processes.

Digital model launch. The digital twin model is being transferred to special platforms, for example, Siemens or Dassault Systemes. The platform integrates mathematical models, data and an interface to control the digital twin, turning it into a dynamic system.

Testing work processes on a digital twin. At this stage, it is predicted how the object or system will behave in normal mode and in emergency situations in order to avoid breakdowns and overload after startup.

Launching and configuring the real object. Some processes are difficult to predict, so they are monitored already at the stage of launching and configuring the real object.

Correction and development of the physical object or system. Specialists continue to collect and analyze data throughout the entire lifecycle of the real object, including using numerous IoT sensors. As a result of this work, changes are made to the physical object to increase its efficiency.

Examples of digital twin applications

ADNOC company has applied digital twin technology to integrate 20 refineries and oil production facilities into a single control room. The digital twin has combined the entire complex of the company’s assets, which are scattered throughout the Middle East. All processes were unified and brought to a single standard. The solution included many functions: predictive analytics, real-time visualization, a system for modeling various scenarios of the enterprise and various incidents, modeling, and optimization of energy consumption.

Schneider Electric plans to implement a digital twin at the Yaya refinery. The digital twin will be used as a digital simulator for operators, a safety system and predictive analytics that will help prevent emergencies. In addition, the project involves the implementation of a cybersecurity system to protect the digital twin of the enterprise from hacking.

Schneider Electric has also completed a pilot project to create a digital twin of its headquarters in Moscow. The project was implemented jointly with EcoDomus within the framework of a memorandum of cooperation in the development of BIM technologies in Russia.

In Sibur company, a digital twin is used to create an engineering data management system, simulate production processes, and simulate production and logistics. In the process of creating a production model, equipment is simulated, data on chemicals and indicators of the technological regime are entered. After checking the model, they perform computational studies and determine the optimal process parameters, search for solutions to improve technological and energy efficiency.

Another digital twin was introduced at KAMAZ, where 3D models of 28 units of CNC machines and 20 universal machines were created, as well as more than 50 units of various technological equipment (robots, manipulators, tilters, roller tables). 3D models are used in modeling machining and assembly, as well as for placing equipment on 3D-layouts of factories.

In Gazprom Neft, pilot implementation of the Digital Field program began in 2014 at the assets of its subsidiary Gazpromneft-Khantos. In 2017, the Production Management Center (PMC) was created, which united all solutions to improve the efficiency of individual production processes of production. One of the key systems of the PMC is a digital twin of the process of lifting fluid from wells, which allows you to select the most optimal operating modes, identify emergency situations in advance, and conduct a preventive assessment of the system’s operation in the event of a change in its configuration. Over time, the MCC will be replenished with other digital twins for systems for maintaining reservoir pressure, power supply, preparation, and utilization of associated gas.

LogistiX company offers a video explaining how to increase warehouse productivity using simulation and optimization modeling (warehouse digital twin)