AeroJEPA: Learning Semantic Latent Representations for Scalable 3D Aerodynamic Field Modeling

TL;DR

AeroJEPA introduces a scalable, latent-space based surrogate model for 3D aerodynamics that improves analysis and design capabilities by decoupling prediction from resolution.

Contribution

It proposes a novel joint-embedding predictive architecture that predicts flow latent representations from geometry and conditions, enabling scalable and semantically meaningful aerodynamic modeling.

Findings

AeroJEPA performs competitively on high-fidelity and large-scale datasets.

The learned latent space encodes meaningful geometric and aerodynamic features.

Latent representations support interpolation, vector arithmetic, and design optimization.

Abstract

Aerodynamic surrogate models are increasingly used to replace repeated high-fidelity CFD evaluations in many-query design settings, but current approaches still face two important limitations: they often scale poorly to the very large fields arising in realistic 3D aerodynamics, and they rarely produce latent representations that are directly useful for analysis and design. We introduce AeroJEPA, a Joint-Embedding Predictive Architecture for aerodynamic field modeling that addresses both issues. Rather than predicting the full flow field directly from geometry, AeroJEPA predicts a target latent representation of the flow from a context latent representation of the geometry and operating conditions, and optionally reconstructs the field through a continuous implicit decoder. This formulation decouples latent prediction from field resolution while encouraging the latent space to organize semantically. We evaluate AeroJEPA on two complementary datasets: HiLiftAeroML, which stresses the method in a high-fidelity regime with extremely large boundary-layer fields, and SuperWing, which tests large-scale generalization and latent-space optimization over a broad family of transonic wings. Across these benchmarks, AeroJEPA is competitive as a continuous surrogate for aerodynamic fields, scales naturally to high-resolution outputs, and learns context and predicted latents that encode geometry and aerodynamic quantities not used directly as supervision. We further show that the resulting latent space supports controlled interpolation, linear probing, concept-vector arithmetic, and a constrained design latent-optimization experiment. These results suggest that predictive latent learning is a promising direction for scalable and design-meaningful aerodynamic surrogate modeling.