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

2026-05-06

AI Summary

This work addresses the limitations of existing aerodynamic surrogate models, which struggle to scale to high-resolution three-dimensional flow fields and lack design-semantic latent representations. The authors propose a novel approach combining a Joint Embedding Predictive Architecture (JEPA) with a continuous implicit neural decoder, enabling resolution-disentangled learning by predicting latent representations of target flow fields from geometric and operational context. This study pioneers predictive latent modeling in aerodynamics, yielding a structured latent space that supports interpolation, linear probing, concept vector arithmetic, and constrained optimization. Experiments on the HiLiftAeroML and SuperWing datasets demonstrate the model’s high-fidelity continuous surrogate capability, scalability to high resolutions, and effective unsupervised encoding of aerodynamic features, highlighting its strong potential for design applications.

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.

Problem

• aerodynamic surrogate modeling
• scalability
• latent representations
• 3D aerodynamic fields
• semantic representation

Innovation

• Joint-Embedding Predictive Architecture
• Semantic Latent Representations
• Scalable Aerodynamic Surrogate
• Continuous Implicit Decoder
• Latent-Space Optimization