Generative AI for e-motor design: a breakthrough showcased at Simulation 2025
The SIA Simulation 2025 Congress, held on April 2–3, 2025, at the Renault Technocentre in Guyancourt, brought together engineers, researchers, and simulation experts to explore the latest innovations in virtual development. Among the highlights of this edition, the paper by Altair Engineering drew strong attention for its pioneering use of Generative AI in electric motor design.
Photo : Aurélien Bacquet
When artificial intelligence designs the motor
Titled "GenAI-based e-motor design", the paper authored by G. Cossio, JP. Berro Ramirez, and other team members like S. Esparza and S. Ghamloush present a study applying diffusion models - a class of generative AI - to the design of electric motors. The goal? To automatically generate rotor geometries optimized for torque performance while respecting manufacturability constraints.
A technical leap presented at Simulation 2025
This work demonstrates how AI can support engineers by generating meaningful initial designs, reducing simulation time, and assisting in multi-objective optimization. By training their models on simulated data using Altair® SimLab® and implementing the model through Altair® AIStudio®, the authors achieved promising results that suggest a future where engineers co-create with AI to tackle complex design challenges.
The congress provided an ideal forum for this type of innovation, emphasizing the strategic role of simulation and AI in accelerating R&D for electrified powertrains, among other domains.
Contact the author :
Juan Pedro Berro RAMIREZ - jpbramirez@altair.com