Observed arrival · 2026-09-18
PINNeAPPle Labs Wants Physics AI to Show Its Work
A physics-AI lab presenting verification tools, neural simulation surrogates, a live river digital twin, and open-source PINN software.
- For
- Engineers evaluating physics-informed AI
- Worth noticing
- The homepage reports 4.46% average and 7.35% worst-case agreement with independent 3D FEA across seven heatsink designs.
Field notes
The project separates its public software foundation from commercial demonstration products. Listed components include PINNeAPPle for PINN, FNO, DeepONet, GNN, solver, and digital-twin work, plus pinnfactory for symbolic PDE definitions using SymPy and automatic differentiation in PyTorch. Its verification example emphasizes per-check reporting: residual, convergence, calibration, and an explicit “not run” state when a check does not execute. A confidential industrial model is cited as an internal guardrail test.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
One card from the complete issue