Aizoth provides "Multi-Sigma-Architect," an AI analysis platform that enables high-speed prediction of shape changes based on design variables and the resulting physical fields, such as stress, flow velocity, temperature, and pressure. Through its proprietary auto-tuning technology, the platform supports the construction of surrogate models with high predictive performance, even when using a relatively small amount of CAE (Computer-Aided Engineering) data. A surrogate model is an approximation model that serves as a substitute for complex simulations.
Product LaunchesAizothMulti-Sigma-Architect
AI analysis platform automatically builds surrogate models from limited data for design optimization
This article is a translation. Read the Japanese original
The platform allows users to perform everything from model construction to analysis via a GUI. During the optimization process, it is possible to search for design candidates that simultaneously satisfy multiple conflicting objectives, such as "maximizing strength and minimizing weight" or "maximizing lift and minimizing drag." Furthermore, it can quantify the contribution of each design variable to performance, which can be used to prioritize design changes.
Analysis results for shapes and physical fields can be output in VTK format, allowing verification using existing visualization or CAD software. Compared to traditional statistical design of experiments (DOE), this approach achieves resolution of multicollinearity issues, improved prediction accuracy, and simplification of experiments.