Siemens limits Physics AI to design exploration, not safety sign-off
Siemens says its Simcenter PhysicsAI software can assess design variations up to 1,000 times faster than a traditional solver. Sam Mahalingam, who leads the relevant business at Siemens Digital Industries Software, said the technology is not suitable for approving safety-critical parts. Its role is to narrow a broad set of options before detailed validation by conventional physics-based simulation.
PhysicsAI uses a surrogate model trained on earlier simulation results. Siemens cites a 1% to 3% difference from physics-based solvers in its case studies, but the company does not regard that level of agreement as sufficient for life-or-death components. Designs selected with the faster system must still pass a full physics-based check before manufacturing.
The model’s performance also depends on its training data. Examples involving Magna and Continental used synthetic data generated with Simcenter HEEDS and Simsolid. Mahalingam acknowledged that the surrogate cannot exceed the quality of the simulations from which it learned. Siemens has added guardrails intended to flag designs that differ substantially from the model’s training range, leaving final validation to engineers.
Written by Sakhanda Wire AI from reporting by AI News. Facts, quotes and full context are in the original article below.