Web-Based Digital Twin
describes only the surface . With the possibility to visualize the data of the linked asset , the webbased digital twin can now digitally represent the state .
The bidirectional connection between the digital twin and the asset is established on the basis of OPC UA information models , which are integrated in an OPC UA server [ 10 ]. Due to the vertical integration of OPC UA and the renunciation of the classic automation pyramid , the OPC UA server can be located anywhere . The OPC UA server contains the status information and provides them to any OPC UA clients . In addition to the simple display of the information units , the state of the asset can be visualized with the help of geometric models .
Another characteristic of a digital twin is the representation of behavior . Representing behavior digitally means replicating how it works through simulations . Performing simulations within the web-based digital twin can be divided into three categories :
• Simulations on the client side
• Simulations on the server side
• Simulations on external servers
The three categories differ in the use of computing resources to perform the simulations . Simulations on the client side use the resources of the requesting client for this purpose . To run the simulations , the desired solution algorithms must be developed in JavaScript . The shift to the client side has advantages if the client has large computing resources and drawbacks if the client does not , e . g . a smartphone . The opposite approach is simulation on the server side . The simulation accesses the same resources as the web server of the web-based digital twin for computation . This also reverses the pros and cons compared to frontend simulation . A third option is to move the computation to external hardware . In this case , the simulation runs on an external server and the results return to the web server . The external server can be self-managed or provided by cloud computing solutions .
The integration of simulations into the web-based digital twin requires the selection of a suitable interface . The multitude of possible interfaces from different simulation environments is a challenge that is addressed by the FMI standard . Currently , more than 100 tools are available for exporting and importing FMUs . With the increasing number of FMUs in the digital twin , the management of the simulation models becomes more important and should be addressed by SDM-DT .
From the corresponding metadata of the models , the respective open simulation parameters with their functional values as well as the output parameters can be read out . The different available simulation models can put together into simulation networks , knowing the input- and output parameters . The use of simulation networks enables time-efficient high-resolution simulations . Crucial for the calculation of the simulations , is the linking of the open simulation parameters with the corresponding sensor values of the asset . Furthermore , discrete simulation
IIC Journal of Innovation - 9 -