RLE develops the future of mobility! From concept design to vehicle studies to series production.
The auto industry is facing market and competitive pressures globally and enhancing product design, development, and a faster time to market are key differentiators.
RLE wants to be at the forefront of AI-driven engineering design for the automotive industry.
RLE developed an end-to-end workflow for CFD predictions using SimScale.
Successfully deployed SimScale-trained ML models for aerodynamic predictions.
Saved 45% on computation costs by using cloud-native simulation with SimScale.
Within seconds, RLE can obtain reliable CFD parameters such as lift, drag, and speed.
This case study examines the innovative approaches adopted by RLE INTERNATIONAL Group in the field of automotive design and testing. By integrating artificial intelligence (AI) tools with cloud-native computational fluid dynamics (CFD) simulations via SimScale, RLE has streamlined the design and testing of automotive components, engines, and systems. The collaboration has successfully enabled RLE to meet the demanding challenges of the automotive industry, including the acceleration of product development cycles and a faster time to market.
RLE INTERNATIONAL Group stands at the forefront of the engineering services industry, particularly in automotive design and testing. With a workforce of over 2,300 employees, RLE leverages cutting-edge technology and innovation to meet its clients’ requirements. The automotive industry faces the daunting task of delivering an increasing number of new products within shrinking timelines and without proportional growth in resources or expenditure. The necessity for cost-effective, rapid, and reliable development of automotive components demands an evolution in engineering processes and tools. To tackle these challenges, RLE has partnered with SimScale, a cloud-native computer-aided engineering (CAE) software, to enhance its capabilities in running CFD simulations and training machine learning models.
RLE aspired to develop a user-friendly end-to-end workflow for CFD prediction tailored to automotive aerodynamics applications. A critical feature of this system is its accessibility to users without extensive prior AI model training, with an optional pathway for detailed training if needed. SimScale’s cloud-based platform facilitates this by allowing the parallel generation of massive datasets, which are pivotal for training localized AI models or for use directly on SimScale’s platform.
For efficiently and independently generating context-specific simulation data, SimScale provided the capability to generate a large number of datasets in parallel, which can be further utilized either in a local AI model or cloud-based on the SimScale platform. Under normal circumstances, license utilization and parallelization on the RLE servers would have been significant costs. This problem does not exist when using SimScale.
Nikola Franic, an AI in engineering expert at RLE, is one of several specialists working with SimScale to develop the new workflow. By leveraging SimScale’s CFD simulation capabilities, RLE has enhanced its engineering workflows, reduced costs, and improved efficiency, positioning itself as a leader in the technological advancement of the automotive industry.
Nikola Franic is an expert in the application of AI in engineering at RLE. He has three years of experience in the automotive industry with a background in mechanical engineering (numerical optimization) and data science.
By using cloud-native CFD in SimScale, the project resulted in savings of 45% compared to using traditional desktop software. This flexibility and cost-saving were the convincing factors for entering into a partnership with SimScale, considering the large scale of customers we serve daily on a global scale.
Nikola Franic
AI Engineer at RLE
“RLE INTERNATIONAL Group is pushing limits in AI prediction for CFD simulation. We have created an end-to-end workflow that includes geometry variation, training data creation using SimScale, model training, model implementation, and subsequent optimization loops using (a third-party tool), Synera! Also, our models are pre-trained for direct use, cloud-based or on-premise.
Tilman Steininger
Technical Unit Lead Connected Engineering at RLE
RLE plans to extend this technology to three-dimensional flow analysis to further increase the quality of design variants. Engineers will be able to significantly reduce the time required to evaluate the influence of hundreds of design variants on vehicle aerodynamics.
Sign up for SimScale
and start simulating now
