RICOS Lightning
RICOS Lightning is a machine learning-driven application designed to predict simulation results. By replacing time-consuming CAE simulations with RICOS Lightning’s predictions, users can significantly reduce computation time and associated costs, including licensing and HPC expenses.
Powered by AI featuring our proprietary IsoGCN and other algorithms, the system learns from 3D geometries in their native form, enabling high-precision predictions for the complex shapes typically encountered in design and engineering workflows.
Scenes
For years, complex geometric simulations have been hindered by long computation times, tedious meshing, and the high cost of evaluating diverse design iterations. RICOS Lightning addresses these challenges simultaneously using machine learning-based predictive models, delivering high-precision analysis results instantaneously.
Acceleratesimulations
CFD simulations for automotive or machinery sectors often demand significant computational time due to the complexity of the phenomena involved.
RICOS Lightning enables rapid prediction of various scalar properties and flow fields in CFD analysis, and we have now expanded our capabilities to include structural analysis.
Performing analysiswith a simplified mesh
By leveraging machine learning, RICOS Lightning achieves high-accuracy results even with coarse meshes that typically compromise traditional simulations.
This enables analysis using simplified models, radically reducing the time and effort required for complex meshing processes.
When analyzing and evaluatingperformance across diverse geometries
To make informed design decisions, performance evaluations across various configurations are essential; however, large-scale simulations are resource-intensive.
RICOS Lightning enables effortless geometry modifications and provides instantaneous analytical results. Please contact us for detailed information or to discuss your specific requirements.
Features
Radical acceleration is achievable
By utilizing machine learning models trained on existing simulation data, we can accelerate computational speeds by over 100 times.
Enabling predictions even for complex geometries
Our proprietary machine learning technology, centered on the Graph Neural Network-based IsoGCN algorithm, enables direct mesh processing, facilitating application to complex geometries (Patent No. 6845364).
Physics-informed machine learning algorithms integrating the characteristics of physical phenomena
Traditional machine learning algorithms often struggle with geometric transformations, viewing rotated or translated shapes as entirely new data, which compromises predictive accuracy.
RICOS Lightning inherently accounts for these spatial shifts, ensuring consistently robust and reliable predictions.
Development status
Currently supported features
- AI prediction of automotive aerodynamic performance
Planned support features
- Thermal conduction analysis prediction AI
- RICOS Optimizer: design optimization
See the Unseen
Please review the materials below for service overviews.
Service
Consulting
Even after introducing our systems, we work alongside customers until the tools are established in practical day-to-day operations. We go beyond simple inquiry handling and collaborate on more effective ways to use the systems based on actual usage and challenges.
Operational Support
Business transformation using AI and advanced technologies cannot be achieved by tools alone. RICOS supports operations after understanding each customer’s workflows and decision-making processes, helping teams continuously improve and create results.