Decision-Aware Evaluation of Physics-Informed Surrogates
For researchers developing physics-informed surrogates, this work provides a reproducible testbed to evaluate models as decision systems rather than just curve predictors, highlighting the gap between curve fidelity and decision utility.
The paper introduces pinn-gym, a benchmark for evaluating physics-informed surrogates in engineering design, showing that low curve error (nRMSE) is insufficient for downstream decision tasks like ranking and feasibility, and that physics-informed losses alter trade-offs across metrics.
Physics-informed machine learning is often assessed by curve error, although engineering use depends on downstream decisions: ranking candidates, avoiding infeasible designs and limiting regret. We introduce pinn-gym, an open benchmark for material-conditioned lattice design that couples a transparent reduced-order crush-and-impact oracle with five printable polymer cards, dimensionless force-response targets and a protocol spanning curve fidelity, physical admissibility, top-k retrieval and mass regret. Across per-material, pooled and cross-material settings, low nRMSE is frequently insufficient to identify useful design selections. Physics-informed losses alter trade-offs rather than monotonically improving all metrics, and dimensionless conditioning improves comparability without making transfer symmetric. The benchmark is not a certified material model; within the released oracle, candidate generator and material cards, pinn-gym provides a reproducible testbed for evaluating PIML surrogates as decision systems rather than curve predictors alone.