AI and System Simulation Seek to Find Better Architecture Structures Faster
Data requirements for engineering projects continue to become more complex, and the engineering world continues to build more sophisticated tools to handle the data they produce. Siemens recently announced Simcenter Studio, a new tool for building system architectures that uses simulation and artificial intelligence (AI).

The nesting structure of the software remains the same but is more complicated. Xcelerator is the top-level tool that uses digital data transfer and enables customers to take highly complex systems and find competitive solutions. Simcenter lives inside and uses digital twin technology to test applications and predict the behavior of systems or objects. Now, Simcenter Studio sits as another application within Simcenter to help engineers develop system architectures.
Current practices for engineers rely on them using past projects as a starting point when designing architectures. The flaws with this approach are obvious: every project is different and will have a different set of requirements and constraints over its life cycle. Projects with tight time constraints or other specialized requirements can be even more sensitive to small (and large) disruptions in the plan. Simcenter Studio aims to remove the current pitfalls associated with continued use of the same framework.

Using the simulation tools, thousands of options are generated for a project. Controllers in the web-based software let the user choose which parameters are important. AI can take the user inputs and identify the best architecture options for a given project. The software uses a Python scripting interface to combine different configurations and produce the file directory structure as an HDF5 file. Although it isn’t shown yet in the material online, a “recommender” tool also exists to show the user options that meet all the criteria but which haven’t yet been viewed or explored.
Studio gives users a computational notebook system that can take in words, equations, codes and models and show the variables all in one place. The tool is complex, and Siemens remains committed to teaching its users through the Xcelerator Academy and the Simcenter Learning Journeys. Even in the promotional material showing the computational notebook, user screens are shown performing development in a main window and using training material in another panel.
Studio is another nice example of Siemens making use of the AI acquisitions the company made over the last year. The commitment and vision are still focused on taking unwieldy complex data points and giving their customers a method to find an optimized path and gain a competitive advantage. There are good points and bad points to adding the functionality as another piece of Simcenter. Customers are required to buy another module for their software, and it adds cost. But it also exists as a tool that users who don’t need it aren’t required to purchase or work around.
Adding the ability to take all the current requirements and constraints to build a data structure in the prelaunch phases of a project is great if used correctly. Engineering often happens on the fly. Team members are sometimes dropped into a project that’s already in process, with many of the decisions already made. Better data planning in the front end of a project will hopefully produce better results for everyone.