How to Make Smart Digital Twins

digital twins dynardoDigital twins continue to grow in importance. Here in Germany, engineers at many companies, including Bosch and Daimler, are dealing with complex applications and the challenge to improve the product performance to come up with an optimized and robust virtual design. They need to determine and evaluate the robustness of virtual prototypes, considering scattering effects, which is difficult or not even possible in hardware tests. Software is used to accurately and rapidly generate proper samples and the resulting understanding saves them a lot of time and money in prototyping so they can stay competitive.

High-performance computing (HPC) and Product Lifecycle Management (PLM) tools are enabling these companies generate and store big data to such an extent that the creation and calibration of virtual system models has become possible. Once created, a digital twin can simulate the product performance at almost real-time operating status. With the resulting diagnoses and prognosis, the performance and reliability of products can be much better understood, controlled and optimized.

For example, engineers can use software to consider measured scatter of various input data in their virtual designs. Then real world data measurements can be input into the digital twin to almost instantly predict the performance and quality of the product.

Using these simulation-based virtual system models, these companies are looking at opportunities for optimization and control of such systems. These tools would enable engineers to speed up the decision-making process, enhance traceability and improve existing Quality Management Systems (QMS) in place. By using these virtual system models to analyze the causes and effects in the operating state, it is also possible to develop intelligent maintenance systems. Digital twins are becoming important components for competitive product development.

Post-processing tools enable full interactive data mining. This is supported through a variety of charts and tables which capture complex statistical evaluations in an easily understandable and presentable way. This capability builds another bridge between the world of hardware data and the world of simulation.

If you would like to learn more about how to apply these methods to help you deliver your product promise, you are invited to our conference. There you can hear engineers from BOSCH, Daimler and other companies who revolutionized their simulation process with these techniques and learn how you can apply CAE-based sensitivity analyses, optimizations and robustness evaluations with ANSYS optiSLang. The 14th Weimar Optimization and Stochastic Days 2017 will take place on June 1-2 offering focused information in practical training and interdisciplinary lectures.