Blog

DIGITAL TWINS IN THE SIEMENS MINDSPHERE IOT SYSTEM

Digital twin and IoT technology is the basis for the successful product development of the future. The goal of the digital twin and IoT is the continuous transmission of data from the operation of each product and their use in the development of each new product. Key players in the market include GE, IBM, Microsoft, Oracle, PTC, Ansys, Dassault Systems and especially Siemens. Digital twin technology removes two barriers for developers working on IoT projects: slowness and complexity. It isolates back-end developers from the complexity of edge-computing development and uses IoT platforms that support digital twin technology and reduce the need for highly specialized device developers.

WHAT IS DIGITAL TWINS FOR?

The digital twin is a detailed digital model that represents products in operating conditions. There are companies that have been using the digital twin for many years. E.g. Apple, where each iPhone records data about its operation and sends it for analysis - so the company obtains data from around the world about which software components work smoothly and which do not. Subsequently, Apple can update the software on each device used, every few weeks. Another representative of the digital twin is, for example, Tesla, which currently creates the digital twin of each vehicle sold. Sensors from thousands of cars continuously transmit data to the numerical model of each car to the manufacturer. Interpretation of data using artificial intelligence (AI) enables fault prediction. Many bugs can be resolved by updating the software. By combining AI and IoT, Tesla can constantly learn from the real world and optimize each individual car in real time.

Virtual prototype vs. digital twin

The terms virtual prototype and digital twin are often referred to as the same thing - this is not the case. The virtual prototype is a numerical model based on FE (finite element) analyzes, flow analyzes (CFD = computational fluid dynamics) and other analyzes using loads for structural design of components. The digital twin also uses a numerical model, but works directly with operating data - voltage, current, temperatures, powers. The digital twin can evaluate risk factors, residual life and monitor the exceeding of design states. The digital twin can also be used for further product optimization. The digital twin can therefore be defined as the interconnection of a real physical product (product) and a digital product (model).

Digital product model

The basis of the digital twin is the product's own numerical model. It can be created in several levels:

- multidomain system 1D model of a complex device

- 3D model based on CFD (computational fluid dynamics), FEM (Finite element model) and MBS (multi-body simulation) technologies

- combined hybrid 1D model with 3D cosimulations or reduced ROM models

1D model

In the case of 1D models, the complex device is divided into individual components, which are replaced by suitable models. Each component requires certain inputs and produces certain outputs. The outputs can then serve as input to another part of the 1D model to create a block diagram. Models are represented by a mathematical function that describes their behavior.

 

The system 1D model is solved for various computational domains representing the physical problems they solve (electrical domain, hydraulic, thermal, mechanical,…). Each of the domains works with different physical quantities.

 

System simulations are used mainly in solving complex systems consisting of many mutually integrating components. For these reasons, the first area of ​​application is the initial dimensioning of new systems. Furthermore, system simulations are used in optimization or what-if analyzes of complex systems (tasks of the type "what happens if I use another instead of this component?"), Or for the design of complex systems control (eg energy optimization of the production process).

 

To create 1D system models, you can use the tool Simcenter Amesim, which contains libraries of predefined models, logic circuits, state diagrams, can work with Matlab, Modelika, Simulink. In the software it is possible to perform cosimulations with 2D / 3D FEM solvers (STAR-CCM +, Fluent, MotorSolve,…), it is possible to edit / create your own models (Visual Basic, C ++, Python, Matlab). The calculation can be performed in both the time and frequency domains and, last but not least, it contains optimization tools that can lead to a rapid improvement of the entire system.

3D model

The 3D model uses physical models and numerical methods based on the finite element method FEM or the finite volume method FVM for solving static and dynamic problems, flow problems, heat transfer and more.

Mindsphere, the IoT system of Siemens

If we have created digital models and can collect data from the operation, it is possible to proceed to the creation of a digital twin. Data is not just an end in itself: it is necessary to analyze it and get the most out of it. And this is where MindSphere comes into play. With MindSphere, you can create a completely closed decision-making environment for continuous optimization.

Key features of the Mindsphere platform include extensive libraries of analytical tools, mapping of physical activities along with virtual models, ensuring secure communication through the Enterprise Gateway, the ability to connect to various PLM systems and simulation tools, and custom connector development tools to connect other business models.

Connecting Mindsphere and Simcenter Amesim

Once we have created a 1D system model, Simcenter Amesim can be connected to operational data directly on the product (eg vehicle) or in the cloud. The system model solver can then be started asynchronously. The Circuit API is used to load simulation packages, run simulations and load simulation results. The results that are computed in the system model are then returned to the Mindsphere platform.

 

Current developments cannot do without the creation of a digital twin. The Mindsphere platform is a solution that allows you to create a completely closed decision-making environment and can lead to continuous optimization of products / processes. The data that is collected can currently be analyzed and used to the maximum (not just for monitoring). The digital twin can evaluate critical states, reliability and longevity and can monitor the exceeding of design states. Digital twin technology brings acceleration and simplification of IoT development solutions. Based on the continuous use of data and constant optimization, the newly developed products are better and the companies that use this technology are market leaders.