Search papers, labs, and topics across Lattice.
DeepJEB++ introduces a novel data-augmentation framework that leverages a foundation model to transform a small seed set of jet engine brackets into a substantial, simulation-labeled 3D dataset. By augmenting in a 2D latent space and then transferring to 3D, the method effectively retains manufacturable designs while automating the labeling process for finite-element analysis. The approach results in a remarkable 40x increase in dataset size, generating 15,360 simulation-labeled 3D brackets from fewer than 400 seed designs, all achieved with minimal computational resources.
A 40x dataset expansion from a mere 400 designs reveals how 2D latent space augmentation can revolutionize 3D engineering data generation.
Data-driven engineering design is constrained by the lack of large-scale 3D datasets that pair geometry with physics-based performance labels. In particular, existing 3D data augmentation techniques have limitations in preserving subtle and diverse geometric variations, and it remains difficult to automate the subsequent simulation-labeling process, where boundary conditions vary depending on the generated geometry. We present DeepJEB++, a foundation-model-driven data-augmentation framework that expands a small seed set of jet engine brackets into a large, simulation-labeled 3D dataset under constrained resources. Our key idea is to augment in the data-rich 2D latent space, then transfer to 3D. In Stage 1, we fine-tune a pretrained 2D latent diffusion model on multi-view renders and synthesize novel views by latent interpolation, retaining manufacturable designs through a vision-language-model (VLM) quality filter. In Stage 2, the validated images are lifted to 3D meshes by a domain-adapted generative foundation model. In Stage 3, an automated pipeline recognizes the load and bolt interfaces on each mesh and assigns finite-element labels -- mass, stress, and displacement -- without manual intervention. We assess augmentation quality along three intrinsic axes: manufacturability, label fidelity against the SimJEB ground truth, and distributional consistency. Starting from fewer than 400 seed designs, DeepJEB++ yields 15,360 simulation-labeled 3D brackets -- a 40x expansion -- using a single GPU per stage. The dataset will be made publicly available to support reproducible engineering-AI research.
