Siemens Digital Industries Software has announced an ongoing collaboration with PhysicsX, an AI startup revolutionizing advanced industries through generative AI technology. The partnership aims to create the next generation of AI-driven deep physics simulations, focusing on aerodynamics to accelerate performance prediction and optimization.
PhysicsX’s latest innovation, a pre-trained deep physics model for aerodynamics, leverages high-fidelity simulation data generated with Siemens' Xcelerator portfolio. This collaboration builds on cutting-edge simulation tools, including Siemens’ Simcenter™ STAR-CCM+™ and Simcenter™ Nastran software, to generate highly accurate data for deep learning applications.
“AI represents a transformative opportunity for the simulation community, and we are exploring its potential at an accelerated rate,” said Jean-Claude Ercolanelli, Senior Vice President of Simulation and Test Solutions at Siemens Digital Industries Software. “Open collaboration is key to scaling AI adoption, and Siemens is excited to see PhysicsX share their progress. Together, we are pushing the boundaries of AI-enhanced simulations that have the potential to fundamentally change product design and engineering.”
Robin Tuluie, Founder and Chairman of PhysicsX, expressed excitement about advancing their collaboration with Siemens. “Successful AI deployment in engineering depends on the highest-quality synthetic data, seamless integration between AI and Computer-Aided Engineering (CAE), and the trust customers place in our technologies,” said Tuluie. “We’re thrilled to announce the release of our Large Geometry Model (LGM) for aerodynamics, LGM-Aero, trained on high-fidelity data from Siemens’ Xcelerator tools. We’re also excited to unveil Ai.rplane, a free-to-use reference application showcasing LGM-Aero’s power in designing innovative aerodynamic solutions.”
Ai.rplane Technology DemonstratorPhysicsX has also launched Ai.rplane, a public-access technology demonstrator designed to provide a set of exploratory tools for aerodynamics and aircraft structures. Ai.rplane uses unique AI-driven shape generation and is powered by LGM-Aero, which is trained on a massive dataset of over 25 million geometries and associated physics simulations. These simulations, which include Computational Fluid Dynamics (CFD) and Finite Element Analysis (FEA), use tens of billions of mesh elements and thousands of simulations from Simcenter™ STAR-CCM+™ and Simcenter™ Nastran.
Ai.rplane represents the world’s first Large Physics Model (LPM) for flight. It is a fully trained model that can generalize across various aeroelastic applications. By using LGM-Aero as a zero-shot model, Ai.rplane can infer aero performance, flight stability, and structural stress for any flying shape generated. In contrast to traditional simulations, which take hours, Ai.rplane creates geometry and assesses physics performance in under a second.
The application allows users to start with basic instructions and quickly explore the generative design space, modifying and optimizing for specific performance characteristics in seconds. Over time, PhysicsX plans to add additional features such as cargo packaging, powertrain selection, and control systems to expand the capabilities of Ai.rplane.
Both LGM-Aero and Ai.rplane are available on the PhysicsX AI engineering platform, enabling users to harness the full potential of AI-driven aerodynamics design and analysis.