close
close
PhysicsX introduces free-to-use “AI for Advanced Engineering” to transform aerospace development

PhysicsX releases the first large-scale geometry model, LGM-Aero, for aerospace engineering, trained and deployed on Amazon Web Services (AWS) cloud computing using more than 25 million different shapes. This reduces the time to investigate aerospace concepts from months to hours.

  • Trained and deployed on AWS, LGM-Aero is a geometry and physics model pre-trained on a corpus of more than 25 million meshes with over 10 billion vertices, enabling dramatic reduction in aircraft concept development time

  • “Ai.rplane” is a publicly available reference application based on LGM-Aero that aims to demonstrate the model’s power to generate novel aircraft designs and predict the physics associated with aircraft performance. It was launched today at AWS re:Invent 2024 (December 2-6), Las Vegas)

  • Both Ai.rplane and LGM-Aero are available on the AI ​​technical platform PhysicsX

  • LGM-Aero’s training data includes a large, high-fidelity physics corpus generated using the Siemens Xcelerator portfolio in collaboration with Siemens Digital Industries

LONDON, December 4, 2024 /PRNewswire/ – PhysicsX, a London-based startup harnessing the power of generative AI to enable breakthrough engineering in advanced industries has launched the first Large Geometry Model (LGM) for aerospace engineering, LGM-Aero, and a publicly available reference application, Ai .rplane, brought to market to showcase its power in designing aircraft structures.

PhysicsX releases the first large geometry model, LGM-Aero, for aerospace engineering
PhysicsX releases the first large geometry model, LGM-Aero, for aerospace engineering

Air.rplane allows engineers to develop innovative aircraft designs in an infinitely large design space and instantly evaluate the potential performance of the designed aircraft.

LGM-Aero was developed and deployed on AWS and trained on more than 25 million meshes representing more than 10 billion vertices, as well as a corpus of tens of thousands of computational fluid dynamics (CFD) and finite element analysis simulations generated with Siemens (FEA). Tools for the digital industry. It is a fully trained model that can be generalized to a wide range of aeroelastic applications. As a zero-shot model, aero performance, flight stability and structural loading are also derived for a large class of flight forms. This technology creates geometry and evaluates performance results in less than a second, compared to the several hours required for traditional numerical simulations.

“We are excited to work with PhysicsX as they develop their first large geometry model and release their flagship application Ai.rplane.” said Ozgur TohumcuGeneral Manager, Automotive and Manufacturing, AWS. “This technology will accelerate the transformation of engineering in advanced industries for AWS customers, enabling them to bring their products to market faster while increasing product performance. We are impressed by PhysicsX’s pace of innovation and look forward to deepening our collaboration.”

Leave a Reply

Your email address will not be published. Required fields are marked *