Digital twin: Fast, yet reliable
How lightning-fast simulations are helping medium-sized companies with million-euro investments
Electricity instead of natural gas, sustainable raw materials instead of petroleum: The chemical industry is undergoing a transformation. However, medium-sized companies often find it difficult to make the right investment decisions for their own business and processes. Prof. Dr. Bastian Etzold, FAU EAM spokesperson and holder of the Chair of Power-to-X Technologies at Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU), is working on a method to simulate different scenarios for an individual company within seconds, providing company leaders with a reliable decision-making tool for costly investments. The project is being funded with 1.3 million euros.
For his project, chemical and bioengineer Prof. Dr. Bastian Etzold is using a digital twin—that is, a copy of the production process created in the digital space, including raw material preparation, chemical conversion, and purification. This digital twin can be used to rapidly simulate company conversions for new sustainable production options and the effects of a wide variety of influences, such as new raw materials or fluctuations in electricity prices.
Why are investments difficult for medium-sized companies?
“Every medium-sized company and every chemical production process is different. Which raw materials are used? Which equipment is used? Which new sustainable raw materials and new technologies can be used? And for a transformation towards sustainability, should existing infrastructures be modified or should completely new production chains be established?” explains Etzold, describing the challenge. “Especially for small and medium-sized enterprises, investment decisions can hardly be tested in a pilot trial and must be made directly. Here, the digital twin helps to try out many different scenarios on the computer. We have been able to run simulations before – but not with this variety of possibilities, nor with this level of precision and speed.”
How does AI help with investments in companies?
Until now, simulations had to be set up individually by people and could only ever represent a specific constellation of parameters – such as the effect of a different raw material or an optimized production process. Bastian Etzold and his team are now working on a method that allows more than 50,000 such computer simulations to be carried out automatically, without a person having to intervene each time to adjust the simulation in case of problems, basically pressing the start button again. This generates large amounts of high-quality data for an immense variety of possible constellations. These data are used to train artificial intelligence (AI). The AI can then quickly predict how a chemical production process will react to different conditions and which factors have which effects. This makes it possible, within seconds, to calculate across a multitude of AI-based simulations how the process can be controlled most efficiently, in the most environmentally friendly way, or at the lowest cost –for example, by using surplus electricity.
About the project
For his funding application in the “European Regional Development Fund: Strategic Technologies for Europe (STEP)” program of the European Union, Etzold succeeded – with the support of the Chemistry Cluster Bavaria – in forming an application consortium: In addition to FAU, eleven medium-sized companies are involved, such as Rösler CeramInno GmbH or OxFA GmbH – they are pilot partners for the project and benefit directly from the research results. The 1.3 million euros in funding will be used, among other things, to finance early career scientists who are working directly on the project topics and who, in addition to the research successes achieved, can also begin their own scientific careers.
Further information:
Prof. Dr. Bastian Etzold
Chair of Power-to-X Technologies
bastian.etzold@fau.de