Proprietary AI That Predicts Simulation Results at Break-Neck Pace: RICOS Lightning
Striving for a world where anyone can use CAE
Make fluid simulation calculations 100x faster —or even more— while maintaining accuracy with our proprietary AI
Recently, the adoption of CAE in manufacturing has been progressing. However, even after implementation, are you still facing challenges in utilizing it effectively, such as the need for manpower, too much cost, and long computation times? Furthermore, the need for coordination and collaboration between departments when using CAE often makes it difficult to fully implement front-loading strategies, doesn’t it?
Front-loading through CAE is particularly critical in product design and development, where frequent prototyping is not feasible.
Our product, powered by proprietary AI, fundamentally accelerates simulations, enabling designers and engineers—without being simulation experts—to utilize it effectively. By integrating this product into your design process, efficiency will significantly improve. Moreover, this enhanced efficiency allows for more simulations to be conducted within the same time frame, leading to higher functionality in products and reductions in development costs.
Significantly streamline simulations
Proprietary AI enables fluid simulations to run 100 times faster.
Among the simulations used for performance evaluation during product design, fluid simulations, in particular, take a considerable amount of time to compute.
RICOS Lightning's key features:
- 100x faster predictions: By utilizing a function learning model trained on existing simulation results, RICOS Lightning can make predictions over 100 times faster than traditional simulations. For example, aerodynamic performance predictions that could take half a day to several days can now be completed in just 10–20 minutes.
- Ideal for predicting complex 3D shapes and objects used in product design: RICOS Lightning's unique machine learning algorithm, IsoGCN, based on graph neural networks, allows for direct handling of mesh data.
- High accuracy on unseen shapes and objects: The algorithm incorporates the characteristics of natural physics, enabling accurate predictions even for unknown shapes and objects. Unlike conventional machine learning algorithms, which struggle with recognizing rotated or translated shapes as the same data, RICOS Lightning correctly interprets such transformations, ensuring highly reliable predictions.
*Depending on conditions, results can vary and be more than 100 times faster.
Intuitive and easy to use
Easy operation allows anyone to predict analysis results effortlessly.
Because of this, it is usable easily by designers and engineers who are not simulation specialists. They can now conduct performance evaluations themselves during the design process, rather than relying on the analysis department as they did in the past. Furthermore, analysts, who often face a high volume of requests, can focus on more specialized tasks by combining AI tools with conventional software to enhance the efficiency of their analytical work.
An example of exploring objects in achieving weight reduction while maintaining product strength.
Enhancing product performance by increasing the design patterns that the AI can evaluate
RICOS Generative CAE's key features:
- Automatic proposal of design modifications within criteria: Our proprietary CAD system automatically generates modified shapes and objects under set constraints and performance evaluations.
- Automatic performance evaluation post-modification: The performance of proposed shape changes is automatically assessed through simulation or AI, supporting methods like fluid and structural analysis.
Automation enables a vast number of trials to be conducted in a short time, leading to cost reduction and improved performance.
- URL (Japanese):https://www.ricos.co.jp/
- URL (English):https://www.ricos.ltd/