About the project
UPSCALE is the first EU project with the specific goal of integrating AI (artificial intelligence) with traditional physics-based Computer Aided Engineering to reduce the development time and increase the performance of electric vehicles (EVs).
Nowadays High-Performance Computing (HPC) and Computer Aided Engineering (CAE) play a decisive role in vehicle development processes, thus the two most HPC and CAE intensive parts of the development, which are vehicle aero-thermal and vehicle crash performance, have been chosen as use cases for the endeavour.
Through the combined effort of universities, research laboratories, European automotive OEMs, software companies and an AI-SME specialized in machine learning (ML), the UPSCALE, project will provide a unique and effective environment to produce novel AI-based CAE-software solutions to improve the competitiveness of the automotive industry.
Discover with this PPT the key aspects of the UPSCALE project!
When considering HPC usage, approximately 20% is being used in aero-thermal simulations and up to 50% is used in crash simulations.
By improving these two areas there will be a 20% reduction of the vehicle development time.
Enhance the performance of existing CFD and FEM crash test tools and processes using machine learning, thus leveraging the potential of ML/MOR to make primary CAE systems faster and more flexible directly impacting the costs and performance of electric vehicles as they heavily rely on simulation for design.
Implement AI for aerodynamic design, first the consortium will work on body parametrization and then will create and train an aerodynamics Reduced Order Models capable of computing aerodynamic values.
Implement AI for design of crash, which will reduce the simulation run time for a full electric vehicle crash by 30%, including battery packs. The consortium will focus on crash battery modelling by means of ROM.
CAE process acceleration by means of subrogation of time-consuming solver functions by AI trained algorithms.
Assessment of a new crash and aerothermal frameworks for full-scale BEV design.