Design & optimization
Surrogate QoI · Solver
- Parametric Exploration / Accelerated DoE
- Multi-Objective Optimization
- Recommending process parameters
A unique combination on the manufacturing market:
Physics + Data + AI
40+ years premium data
600+ partners across 30+ countries
Mines Paris-PSL scientific backing
Production & control
Surrogate QoI
- Real-time process control and correction
- QoI monitoring (ovality, flatness, etc.)
- Digital Process Twins
Materials & acceleration
Surrogate Physics · Solver
- Learned laws of behavior (rheology, hardness, heat transfer)
- Calibration on experimental data
- Fast full fields, lower FEM cost
"Industrial AI is nothing without the quality of the simulations it feeds on."
Surrogate Models for Manufacturing
technology at a glance
Three families of surrogate models, from development to deployment — covering design, production, and data exploitation.
01 / 03
Surrogate Physics
Local Physics
Replaces a local physics law (behavior, transfers, ..) with a data-learned model, embedded in the solver.
Deployed
- Data Mining
- Experimental Calibration
02 / 03
Surrogate QoI
Black Box Predictors
Directly predicts quantities of interest (QoI) without a full field — fast.
Industrial PoC
- Design optimization
- Production control
- Quality monitoring
03 / 03
Surrogate Solver
Full fields
Solver substitute producing full fields via GNN-FEM hybrid architectures.
In development
- Digital Twins
- Quicker simulation fine-tuning
Cycle covered
01
Design →
02
Manufacturing / Control →
03
Material & Data Exploitation
Two directions of use
forecasting
direct problem
optimization
inverse problem
Two directions of use: forecasting (direct problem) and optimization (inverse problem).
Contact Us
JOSE ALVES
Andres RODRIGUEZ-VILLA