DeepWheel: Generating a 3D Synthetic Wheel Dataset for Design and Performance Evaluation

2025-04-15

AI Summary

A critical bottleneck in automotive wheel hub design is the lack of large-scale, high-fidelity 3D geometric–physical performance paired datasets. Method: This paper introduces the first multimodal synthetic data generation paradigm tailored for engineering design. It proposes a closed-loop framework integrating Stable Diffusion for conditional image generation, MiDaS-driven 2.5D reconstruction, SIMP-based topology optimization, and ANSYS structural simulation feedback—enabling geometry-controllable, physics-verifiable inverse design and systematic exploration of the design space. Contribution/Results: We release DeepWheel, the first open-source benchmark dataset for data-driven intelligent design (CC BY-NC 4.0), comprising over 6,000 high-fidelity multi-view 2D images and 900 parametric 3D wheel hub models, each annotated with comprehensive mechanical performance labels (stress, displacement, mass).

Technology Category

• Search and Optimization: Sampling/Simulation-based Search
• Computer Vision: Diffusion Models for Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Web Mining and Content Analysis: Web data generation and simulation
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Data-driven design is emerging as a powerful strategy to accelerate engineering innovation. However, its application to vehicle wheel design remains limited due to the lack of large-scale, high-quality datasets that include 3D geometry and physical performance metrics. To address this gap, this study proposes a synthetic design-performance dataset generation framework using generative AI. The proposed framework first generates 2D rendered images using Stable Diffusion, and then reconstructs the 3D geometry through 2.5D depth estimation. Structural simulations are subsequently performed to extract engineering performance data. To further expand the design and performance space, topology optimization is applied, enabling the generation of a more diverse set of wheel designs. The final dataset, named DeepWheel, consists of over 6,000 photo-realistic images and 900 structurally analyzed 3D models. This multi-modal dataset serves as a valuable resource for surrogate model training, data-driven inverse design, and design space exploration. The proposed methodology is also applicable to other complex design domains. The dataset is released under the Creative Commons Attribution-NonCommercial 4.0 International(CC BY-NC 4.0) and is available on the https://www.smartdesignlab.org/datasets

Problem

• Lack of large-scale 3D wheel datasets with performance metrics
• Need for synthetic data generation for wheel design innovation
• Limited resources for data-driven inverse design and exploration

Innovation

• Generative AI creates synthetic wheel dataset
• 2D to 3D conversion via depth estimation
• Topology optimization diversifies design space